Integrated Object-Based Image Analysis for semi-automated geological lineament detection in southwest England

Accepted Manuscript Integrated Object-Based Image Analysis for semi-automated geological lineament detection in Southwest England Christopher M. Yeomans, Maarit Middleton, Robin K. Shail, Stephen Grebby, Paul A.J. Lusty PII:

S0098-3004(18)30303-0

DOI:

https://doi.org/10.1016/j.cageo.2018.11.005

Reference:

CAGEO 4192

To appear in:

Computers and Geosciences

Received Date: 21 March 2018 Revised Date:

9 September 2018

Accepted Date: 9 November 2018

Please cite this article as: Yeomans, C.M., Middleton, M., Shail, R.K., Grebby, S., Lusty, P.A.J., Integrated Object-Based Image Analysis for semi-automated geological lineament detection in Southwest England, Computers and Geosciences (2018), doi: https://doi.org/10.1016/ j.cageo.2018.11.005. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

RI PT

Integrated Object-Based Image Analysis for semi-automated geological lineament detection in Southwest England

Christopher M. Yeomansa,b,1,∗, Maarit Middletonc,2 , Robin K. Shaila,3 , Stephen Grebbyd,3 , Paul A. J. Lustyb,4 a

M AN U

SC

Camborne School of Mines, College of Engineering, Maths and Physical Sciences (CEMPS), University of Exeter, Penryn Campus, Penryn, Cornwall, TR10 9FE b British Geological Survey, Environmental Science Centre, Keyworth, Nottinghamshire, NG12 5GG c Geological Survey of Finland, P.O. Box 77, FI-96101 Rovaniemi, Finland d University of Nottingham, Nottingham Geospatial Institute, Innovation Park, Nottingham, NG7 2TU

Abstract

AC C

EP

TE D

Regional lineament detection for mapping of geological structure can provide crucial information for mineral exploration. Manual methods of lineament detection are time consuming, subjective and unreliable. The use of semiautomated methods reduces the subjectivity through applying a standardised method of searching. Object-Based Image Analysis (OBIA) has become a mainstream technique for classification of landcover, however, the use of OBIA methods for lineament detection is still relatively under-utilised. The Southwest England region is covered by high-resolution airborne geophysics and LiDAR data that provide an excellent opportunity to demonstrate the power of OBIA methods for lineament detection. Herein, two complemen∗

Corresponding author Email address: [email protected] (Christopher M. Yeomans) 1 Corresponding author, PhD student conducting research, wrote and compiled manuscript 2 Collaborator from partner institution, provided time, training and facilities whilst CMY visited offices, significant contributions to reviews of manuscript 3 PhD supervisor to CMY, providing academic support and reviews on manuscript 4 PhD supervisor to CMY, won grant for PhD funding, provides support from nonacademic institution (BGS) to CMY

Preprint submitted to Computers & Geosciences

November 13, 2018

ACCEPTED MANUSCRIPT

M AN U

SC

RI PT

tary but stand-alone OBIA methods for lineament detection are presented which both enable semi-automatic regional lineament mapping. Furthermore, these methods have been developed to integrate multiple datasets to create a composite lineament network. The top-down method uses threshold segmentation and sub-levels to create objects, whereas the bottom-up method segments the whole image before merging objects and refining these through a border assessment. Overall lineament lengths are longest when using the top-down method which also provides detailed metadata on the source dataset of the lineament. The bottom-up method is more objective and computationally efficient and only requires user knowledge to classify lineaments into major and minor groups. Both OBIA methods create a similar network of lineaments indicating that semi-automatic techniques are robust and consistent. The integration of multiple datasets from different types of spatial data to create a comprehensive, composite lineament network is an important development and demonstrates the suitability of OBIA methods for enhancing lineament detection.

3 4 5 6 7 8 9 10 11 12 13 14 15

Mapping of geological structures can be both time-consuming and challenging in the field, particularly in areas of poor outcrop exposure. Structures such as strike-extensive faults can be particularly difficult to map by a geologist in the field due to partial exposure and subtle topographic variations. Lineament detection can aid the mapping of geological structure. A lineament is a mappable rectilinear or curvilinear linear feature of a surface, distinct from adjacent patterns, and which may represent a subsurface phenomenon (O’Leary et al., 1976). Remotely sensed data, including satellite imagery and airborne geophysical data, are commonly used for regional lineament mapping. These lineaments can be used to infer the structural geology of the region and have implications for mineral exploration (Moore and Camm, 1982; James and Moore, 1985; Ni et al., 2016), petroleum exploration (Pe˜ na and Abdelsalam, 2006), groundwater studies (Kresic, 1995; Mallast et al., 2011) and natural hazard assessment (Rutzinger et al., 2007).

EP

2

1. Introduction

AC C

1

TE D

Keywords: Object-based Image analysis, airborne geophysics, lineament analysis, semi-automated, integration

16

2

ACCEPTED MANUSCRIPT

19 20 21 22 23 24 25 26

RI PT

18

Classical lineament extraction techniques include manually digitising linear features. At the regional scale, optical imagery is commonly used, however, this is time consuming and subjective and therefore lacks reproducibility (Masoud and Koike, 2006; Scheiber et al., 2015). Grebby et al. (2012) demonstrated that airborne Light Detection and Ranging (LiDAR) data can have advantages over optical data where obstructions caused by vegetation cover can be removed to produce a Digital Terrain Model (DTM). Whilst somewhat subjective, human bias when manually digitising can be mitigated by careful data assessment (Scheiber et al., 2015), although this does not make it a less time-consuming approach.

SC

17

27

32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54

M AN U

31

TE D

30

The automation of geological lineament mapping from remotely sensed data has been a key research topic in structural geology for some decades. Automation overcomes issues of time and subjectivity that affects the manual approach. Whilst fully automatic methods are beyond the scope of current computer algorithms, semi-automated techniques are improving rapidly. Early criticisms of semi-automatic methods included the misidentification of roads and field boundaries as lineaments as well as the adverse effect of vegetation cover in some environments. These can now be circumvented with more sophisticated processing and the use of non-optical datasets such as airborne geophysics, especially magnetic data (Middleton et al., 2015), electromagnetic data (Paananen, 2013), radiometric data (Debeglia et al., 2006) and gravity data (Lahti et al., 2014). Semi-automated mapping often makes use of an image enhancement step which can include a transform or filter. Transforms convert measurable parameters to different measurable parameters, whereas filters selectively remove undesirable data at particular frequencies (Milligan and Gunn, 1997). Enhancement filters include convolution, directional, Laplacian and Sobel filters and Principal Component Analysis (Sukumar et al., 2014). Transforms used for lineament detection can search for edges (sharp gradients) in an image (e.g. Blakely and Simpson, 1986) or the presence of elongated minima or maxima in the data (e.g. Airo and Wennerstr¨om, 2010). Transforms are more common when using potential field data such as airborne magnetic and gravity data. These often use horizontal or vertical derivatives to enhance geological edges. An effective option is the Tilt Derivative (TDR), which acts as a gain control, normalizes and enhances the continuity of structures (Verduzco et al., 2004) to help detect edges, minima or maxima (Fairhead

EP

29

AC C

28

3

ACCEPTED MANUSCRIPT

56

et al., 2004; Airo and Wennerstr¨om, 2010; White and Beamish, 2011; Middleton et al., 2015).

RI PT

55

57

60 61 62 63 64 65 66

Lineament detection algorithms vary considerably in approach. The earliest of these were developed for use with Landsat TM satellite imagery such as Edge Following and Graph Searching developed by Wang and Howarth (1990) and the Segment Tracing Algorithm (STA) by Koike et al. (1995). Algorithms used with potential field data, such as those by Lee et al. (2012) ˇ and Silhav´ y et al. (2016), employ curvature and hillshade enhancements, respectively. Most recently, the LINDA software has been developed as a comprehensive tool for lineament detection for spectral and potential field data, and incorporates the STA algorithm Masoud and Koike (2017).

SC

59

M AN U

58

67

71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92

TE D

70

Object-Based Image Analysis (OBIA) is a powerful tool for analyzing spatially correlated groups of pixels. It utilises image segmentation algorithms to group contiguous pixels into image objects. The advantages of image objects are that the grouped pixels can be assessed together to measure texture and geometry and have corresponding summary statistics, whilst being linked geospatially through a topology (Lang, 2008). The generation of representative image objects is of paramount importance to OBIA and therefore some user knowledge is required for a successful analysis (Blaschke et al., 2004). The idea of user knowledge refining image objects is considered to be the change from an object-based workflow to an object-oriented workflow (Baatz et al., 2008). For simplicity, we retain the term object-based in this paper. Despite the adoption of OBIA techniques within the geosciences in the early 2000s, geological studies using OBIA techniques are largely restricted to classification studies (e.g. Grebby et al., 2016) and not to lineament detection. Previous studies for using an OBIA for lineament detection include Marpu et al. (2008), Mavrantza and Argialas (2008) and Rutzinger et al. (2007), who applied the method to Synthetic Aperture Radar (SAR), Landsat and LiDAR DTM data, respectively. The first application of the OBIA techniques to airborne geophysics was reported by Middleton et al. (2015) who utilised airborne magnetic data over the Enonteki area of northern Finland.

EP

69

AC C

68

The objective of this study is to develop semi-automated OBIA methods for geological lineament detection that integrate airborne geophysical and remote sensing datasets to create a composite lineament network. The topdown OBIA method from Middleton et al. (2015) is developed to integrate 4

ACCEPTED MANUSCRIPT

99 100 101 102 103 104 105 106 107 108 109 110 111 112 113

114 115 116 117 118 119 120 121 122 123 124 125 126 127 128

RI PT

98

1.1. The Southwest England region The Tellus South West project (www.tellusgb.ac.uk) flew high-resolution regional airborne surveys over SW England, the coverage of which is outlined in Figure 1. The geology of SW England provides an excellent opportunity to test OBIA methods to detect lineaments in a region where the structural geology is well-documented at multiple scales. The temperate climate of SW England limits well-exposed areas to coastal wave-cut platforms. Therefore, detailed structural studies are focussed along coastal sections and accurate field mapping inland is problematic due to soil cover. The high-resolution regional airborne geophysical surveys provide an excellent dataset for detection of geological lineaments. SW England is one of the most prospective areas within the UK for both metalliferous and industrial minerals and deep geothermal energy. The occurrence and exploitation of these resources are strongly influenced by regional fault networks. Detailed lineaments maps will help inform future exploration activity.

SC

97

M AN U

96

TE D

95

1.1.1. Previous Lineament Studies in SW England Early attempts at regional lineament detection in SW England were conducted on Landsat MSS data (Moore and Camm, 1982). The study aimed to identify major structural lineaments relating to tin-tungsten mineralisation. Due to the coarse resolution of the imagery, the observations were only appropriate for 1:250 000 scale mapping, although significant NW-SE structures were be clearly identifiable. A further study was conducted by James and Moore (1985) to investigate the temporal variations between different scenes and the incorporation of eastward-looking SAR data from Seasat imagery to enhance structural lineaments - although spatial resolution remained low.

EP

94

multiple datasets and compared to a new approach presented here which uses a bottom-up OBIA approach. The Southwest (SW) England region is used as a case study due to the complex structural geology and the availability of high-resolution airborne geophysical datasets. Final lineament maps are contrasted with a regional fault map from 1:50 000 British Geological Survey mapping districts in the study area.

AC C

93

Regional lineament detection using Landsat TM satellite imagery was first attempted by Smithurst (1990) and demonstrated good agreement with mapped major structures. A more detailed study by Rogers (1997) used two Landsat TM scenes which were reprocessed to 150 m pixels and applied four 5

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

130 131 132 133 134

directional filters (N, E, NE, SE). Lineaments were manually digitised based on four criteria: a width/length ratio > 0.5; within ±30◦ of the directional filter; be larger than four pixels; a qualitative strike length significance (Rogers, 1997). The study was effective but was hindered by cloud cover and rejected lineaments < 4 pixels long, therefore neglecting significant structures < 600 m in length.

AC C

129

EP

TE D

Figure 1: Regional geology of SW England depicting the Upper Palaeozoic sedimentary basins and Permian Granites (N.B. the Upper Devonian South Devon and Tavy basins are merged for simplicity). The black box delimits the Tellus South West airborne geophysical data and the rectangular enclosure represents the inset over the Tamar Estuary at Her Majesty’s Naval Base Devonport. C = Cenozoic; PS = Permian Sandstones; CBH = Culm Basin (Holsworthy Group); CBTV = Culm Basin (Teign Valley Group); NDB = North Devon Basin; GB = Gramscatho Basin; SDBTB = South Devon Basin and Tavy Basin; LB = Looe Basin; LC = Lizard Complex; SC = Start Complex; PG = Permian Granites. c Crown Copyright and database right (2018). This figure includes OS data

135 136 137

A more recent study mapping strike-slip faults along the north Cornwall coast at Westward Ho! demonstrated that, in well-exposed areas, rapid 6

ACCEPTED MANUSCRIPT

141 142 143 144 145 146 147 148 149 150 151 152

RI PT

140

1.1.2. Geological Setting The bedrock geology of SW England comprises Devonian-Carboniferous sedimentary rocks, metamorphosed to sub-greenschist facies during the Variscan Orogeny and intruded by a Permian granite batholith (Figure 1). The sedimentary successions were deposited in six E-W-trending extensional basins which were subsequently inverted during the Variscan Orogeny (Leveridge and Hartley, 2006). The Variscan Orogeny resulted in two phases of NNWdirected thrusting which formed an ENE-WSW to E-W-trending fault system (Alexander and Shail, 1995, 1996; Leveridge et al., 2002; Leveridge and Hartley, 2006). Steeply-dipping NW-SE and subordinate NNE-SSW strike-slip faults (Dearman, 1963) developed during Variscan convergence (Leveridge et al., 2002); some NW-SE faults may have been inherited from the preVariscan extensional regime (Roberts et al., 1993; Shail and Leveridge, 2009).

SC

139

structural mapping from aerial photography is a valuable approach (Nixon et al., 2011).

M AN U

138

153

157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174

TE D

156

EP

155

The batholith generated and emplaced in an Early Permian post-orogenic extensional setting and the exposed plutons have a spatial correlation with major NW-SE faults (Dearman, 1963; Alexander and Shail, 1996). The extensional reactivation of ENE-WSW to E-W striking Varsican thrusts was accompanied by the creation of new steeply-dipping conjugate extensional faults, that often host granite-related W-Sn-Cu mineralisation, and was followed by minor intra-Permian shortening episodes (Shail and Alexander, 1997). Subsequent latest Permian to Triassic ENE-WSW extension reactivated older NW-SE faults and created new NW-SE to N-S faults, “cross-courses”, that cut earlier granite-related mineralisation and locally host epithermal basinal brine mineralisation (Scrivener et al., 1994; Shail and Alexander, 1997). The regional fracture network had essentailly formed by the late Triassic. However, several major NW-SE faults in the east of the region underwent Oligocene sinistral reactivation to form pull-apart basins (Holloway and Chadwick, 1986; Gayer and Cornford, 1992) and their infil is locally deformed in a presumed Miocene dextral strike-slip regime (Holloway and Chadwick, 1986).

AC C

154

Whilst the structural evolution, and development of regional fault networks in SW England is complex, the dominant structural trends are ENEWSW to E-W and NW-SE with subordinate NE-SW and N-S elements. 7

ACCEPTED MANUSCRIPT

181 182 183 184

185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210

RI PT

180

SC

179

2.1. Airborne Geophysical Datasets Data from the Tellus South West Project provide a new opportunity to accurately determine lineaments at high resolution across SW England. The airborne data were acquired in two surveys; the LiDAR data collected in summer 2013 and the magnetic and radiometric data in late 2013 to early 2014 (Beamish and White, 2014). The surveys have near-complete coverage, except for an inset of data within the magnetic and radiometric data which was not collected over the Tamar Estuary at Her Majesty’s Naval Base (HMNB) Devonport (Figure 1).

M AN U

178

The airborne geophysical data used in this study were selected to demonstrate the strength of the method for integrating multiple datasets to produce a composite lineament network. The methods herein describe the preprocessing of the data before analysis in eCogntion software (v9.3, Trimble, Germany). Post-processing was also necessary for quality assessment of the data and to create and manipulate metadata. Full details of the software used in these steps can be found in the Supplementary Information (S1). The two OBIA methods presented here can be downloaded individually as an eCognition Rule Set from the University of Exeter at empslocal.ex.ac.uk/obld2.

TE D

177

The magnetic and radiometric data were collected simultaneously by CGG Airborne Survey (Pty). Final quality assurance and quality control of the data were undertaken by the British Geological Survey (Beamish and White, 2014) and the Centre for Ecology and Hydrology (Gerard, 2014). The data were gridded using a minimum-curvature algorithm and supplied at a pixel resolution of 40 m. The survey parameters are summarised in Table 1. In this study, the Total Magnetic Intensity (TMI) from the magnetic data (corrected using the International Geomagnetic Reference Field) and the Total Count channel from the radiometric data were used.

EP

176

2. Materials and Methods

AC C

175

The LiDAR data were collected as part of a separate survey at 1 point/metre2 with 25 cm vertical accuracy. The data were acquired by the British Antarctic Survey, processed by Geomatics (Environment Agency), and overseen by the Centre for Ecology and Hydrology (CEH). Some areas of low quality data acquisition exist in urban areas and dense vegetation (Gerard, 2014). The data were supplied in raster format as a DTM with a pixel resolution of 1 8

ACCEPTED MANUSCRIPT

211 212

m, and are used here under an Open Government Licence (Ferraccioli et al., 2014).

213

217 218 219 220 221 222

223 224 225 226 227 228 229 230 231

232 233 234

TE D

216

The initial airborne geophysical data from the Tellus South West project are illustrated in Figure 2A-C. The three datasets each add different information to the analysis. The LiDAR data captures lineaments that are manifest in the geomorphology of the region. Lineaments derived from magnetic data may indicate the presence of geological structure extending at depth in the crust. The radiometric data, specifically the Total Count channel used here, senses lineaments that may have acted as fluid conduits resulting in mineral alteration. Lower values therefore represent the leaching of radiogenic elements such as potassium and uranium. 2.2. Lineament Detection The two different OBIA methods described here represent a novel approach to geological lineament detection where airborne geophysical and remote sensing datasets can be integrated in order to produce a comprehensive and composite lineament network. The methods primarily use the eCognition software (v9.3, Trimble, Germany) and the Cognitive Network Language (CNL). The CNL provides a variety of tools for OBIA workflows. The workflow for the two stand-alone OBIA methodologies using top-down and bottom-up segmentation techniques is summarised in Figure 3.

EP

215

AC C

214

LiDAR 700 m E-W and ENE-WSW N/A 1100 m Unknown 1 point/m2

SC

Magnetic Radiometric 200 m 200 m N-S N-S 2000 m 2000 m 91.55 m 91.55 m 71.21 m/s 71.21 m/s 20 Hz 1 Hz

M AN U

Survey Parameter Line Spacing Line Orientation Tie Lines Average Altitude Speed Sample Frequency

RI PT

Table 1: Survey parameters for the Tellus South West airborne geophysical survey compiled from Beamish and White (2014) and Gerard (2014). The separate LiDAR survey flew an ENE-WSW angle to align with the axis of the peninsula to the west.

2.2.1. Pre-processing Initial pre-processing was conducted in R project software (https://cran.rproject.org) using the raster package (v2.6-7, Hijmans et al., 2017). All 9

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

10 Figure 2: Input data from Tellus South West (A) magnetic data in TMI; (B) radiometric data in Total Count; (C) LiDAR DTM data in metres; (D) TDR-transformed magnetic data; (E) TDR-transformed radiometric data; (F) TDR-transformed LiDAR DTM data.

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Figure 3: Workflow for creating a full lineament network using both OBIA methodologies. The split in the flowchart indicates the slightly different path for each method rather than parallelised analysis.

238 239 240 241 242 243 244 245 246 247 248 249 250 251

TE D

237

datasets were resampled to the same extent and resolution (40 m pixels) using a bidirectional operator before being clipped to the coastline. The dashed rectangular area missing in the magnetic and radiometric data over HMNB Devonport was not removed from the LiDAR data (Figure 1). The magnetic and LiDAR data underwent further preprocessing steps in the Oasis Montaj programme (v8.5, Geosoft, Canada) to remove noise. These include cultural artefacts relating to buildings and infrastructure as well as minor corrugations across flight lines. A smoothing algorithm was applied in Oasis Montaj to mitigate the artefacts from densely vegetated and urbanised areas in the LiDAR DTM. As a result, the topographic expression of anthropogenic features such as road cuttings were drastically reduced and smoothed. The smoothed product was carefully compared to the original data to ensure minimal loss of geological information.

EP

236

AC C

235

The final pre-processing step applies the TDR function to enhance the continuity of the data (Equation 1).

11

ACCEPTED MANUSCRIPT





RI PT

∂T     −1  ∂z  T DR = tan  s   2  2  ∂T ∂T  + ∂x ∂y 252 253

256 257 258 259 260 261

The TDR transform uses the arctangent of the ratio of vertical derivative to total horizontal derivative and therefore normalizes the data to the range π/2 to −π/2 (Miller and Singh, 1994). The normalization allows comparison of the data and the ratio of the derivatives acts as a gain control to ensure minor gradient changes are not lost (Verduzco et al., 2004). The TDR transform is applied to all three datasets, illustrated in Figure 2D-F. For the DTM and Total Count data, the vertical derivative is calculated using convolution rather than Fourier transform.

SC

255

M AN U

254

262

266 267 268 269 270 271 272 273 274

275 276 277 278 279 280 281 282

TE D

265

EP

264

The TDR transform is commonly used to identify the edges of isolated magnetic bodies at the zero contour of the transformed data (Fairhead et al., 2004; Beamish and White, 2011; White and Beamish, 2011). The TDR transform here is applied here to search for minima; these are considered an effective proxy for lineaments related to fault zones. It is assumed the minima are generated through preferential erosion (LiDAR DTM), breakdown of magnetic minerals (magnetic data) or leaching of radioelements (radiometric data). It is acknowledged that in some geological environments lineaments may be represented by maxima (indurated rocks or magnetic and radiometric highs). Some knowledge of the geology is key to deciding whether to search for minima or maxima, or both. Full details of the pre-processing steps can be found in the Supplementary Information (S2).

AC C

263

(1)

2.2.2. Line Extraction The line extraction algorithm in the CNL was used to create a newly derived lineness raster from each input dataset. The algorithm uses a userdefined, rectangular kernel filter to assess the similarity of pixels. The kernel was set to search at regular interval angles of 5◦ . The parameters for line extraction were optimised to be the same across all datasets to produce three lineness rasters with similar attributes that could be subsequently integrated. The resultant raster has a range of 0-255; higher values indicate a higher 12

ACCEPTED MANUSCRIPT

285 286 287 288 289 290 291

RI PT

284

probability that the lineament exists. A detailed description of the algorithm within eCognition is provided in the Supplementary Information (S1). 2.2.3. Top-down Segmentation A top-down methodology involves the segmentation of an image into smaller and smaller objects and the downward propagation of object levels in the CNL (Diamant, 2004). Determining object relationships across different levels such as super-objects is advantageous for manipulating specific parts of the original object (such as the end) and is used in this instance to preferentially grow objects at their ends.

SC

283

292

294 295 296 297 298 299

Lineament detection through a top-down segmentation was achieved using the multi-threshold segmentation and chessboard segmentation tools in the CNL. The workflow is based on the Object-Based Lineament Detection (OBLD) algorithm of Middleton et al. (2015), which is further developed here to integrate all three input datasets to create a composite lineament network. The OBLD algorithm is discussed in full within Middleton et al. (2015).

M AN U

293

300

304 305 306 307 308 309 310 311 312 313 314 315 316 317

TE D

303

The integration steps developed here for the top-down OBIA method involve the merging of separate image object sets from each dataset where objects were optimised for all three datasets. A complementary merged lineness raster is created by summing the individual lineness rasters and normalising to 0-255. Objects are extended in a similar fashion using a new sub-level under the existing data, as is the case in Middleton et al. (2015), and further segmented using the chessboard segmentation tool. However, in this step, extensions are determined from the merged lineness raster and therefore could be based on values derived from a different dataset. An additional requirement is that object-ends are only extended if in proximity to another object-end. A final cleaning step removes any spurious objects that are below a given area or asymmetry threshold.

EP

302

The general workflow for lineament detection using top-down OBIA methods from multiple datasets is summarised in Figure 3. The segmentation steps and use of object-levels for extending objects are illustrated in Figure 4A1-4.

AC C

301

13

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Figure 4: Segmentation methods for each OBIA approach. (A) illustrates the top-down method where 1 = the lineness raster; 2 = multi-theshold segmentation; 3 = extension through sub-levels, chessboard segementation and super-objects; 4 = object merging. (B) illustrates the bottom-up method where 1 = the lineness raster; 2 = multiresolution segmentation; 3 = spectral difference merging; 4 = border assessment; 5 = object merging.

14

ACCEPTED MANUSCRIPT

319 320 321 322 323

2.2.4. Bottom-up Segmentation The bottom-up method for lineament detection involves the segmentation of the whole image, using the multi-resolution segmentation tool, into many, differently sized, objects which are subsequently merged (Dragut et al., 2010; Eisank et al., 2014). Merging can be based on spectral, statistical, textural or geometric properties and is highly versatile.

RI PT

318

324

327 328 329 330 331 332 333

SC

326

The integration step occurs prior to segmentation in the bottom-up method. As before, a composite lineness raster is created by summing the lineness rasters and normalising to 0-255. Figure 4B1-5 illustrates how the composite lineness raster is segmented using the multi-resolution segmentation algorithm and the subsequent merging of objects using the spectral difference algorithm. Next, the resulting objects are then classified based on the mean pixel value in the lineness raster into major and minor lineaments according to a threshold defined by the user. The threshold was defined heuristically based on user knowledge of the region, and is thus case-specific.

M AN U

325

334

338 339 340 341 342 343 344 345 346 347 348 349 350

351 352 353 354

TE D

337

Major and minor lineaments are refined using a border assessment to merge segments of a lineament which may contain both major and minor components into a continuous object set of either major or minor objects. Initially, major lineament objects are expanded by one pixel to minimise single pixel borders with minor lineament objects. Minor lineament objects which have the majority (60%) of their relative border in contact with major objects are converted into major lineament objects. Major lineament objects are shrunk by growing unclassified objects into the major object where the composite lineness raster is below the heuristically defined threshold for a major lineament. The remaining major lineament objects are then merged. Major objects are converted into minor objects where the majority (60%) of the relative border is in contact with a minor lineament. Cleaning steps remove objects with an area below a given threshold (in this case 50 pixels for major, and 30 pixels for minor lineaments) and also remove pixels at the edges of objects below a threshold in the composite lineness raster; this is tailored to major and minor objects.

EP

336

AC C

335

2.2.5. Vectorization and Metadata The final step using the CNL was to convert objects to polylines and export these to a shapefile with the correpsonding metadata. The polylines were vectorized into skeleton polylines of the objects they represent. Skeleton 15

ACCEPTED MANUSCRIPT

356

polylines are most efficient at capturing objects that represents a complex lineament network.

RI PT

355

357

360 361 362 363

364 365 366 367 368

The metadata for the top-down segmentation method includes binary fields detailing the source dataset for each lineament (extension, LiDAR DTM, magnetic, radiometric). The end-user can manipulate these fields to fully describe the source data for the lineament. For the bottom-up segmentation method, metadata was restricted to the classes of either major or minor lineaments.

SC

359

2.2.6. Post-processing Post-processing of the metadata involved the calculation of the orientation and length of polylines and was conducted in the ESRI ArcGIS software. The computed polyline length, orientation and a direction (N, NE, E, or SE) were included in the metadata of the vector file.

M AN U

358

369

377

3. Results

373 374 375

378 379 380 381 382

383 384 385 386 387 388

The polyline data for both top-down and bottom-up methods were assessed for line length and orientation. A qualitative analysis of the data is presented as well as comparison with the existing regional map compiled from the 1:50 000 DiGMap data from the British Geological Survey. The polyline output for the top-down and bottom-up methods are presented in Figure 5.

EP

372

AC C

371

TE D

376

The polyline data for both top-down and bottom-up methods were interrogated through database queries in the GIS. The data from the top-down method were found to have 29 artefact polylines associated with the extension step of the method. These artefacts, defined as disconnected polylines with lengths >40 metres, were removed from the final lineament dataset. Full details of the post-processing steps can be found in the Supplementary Information (S3).

370

3.1. Statistical Analysis Summary statistics for each polyline dataset are presented in Table 2 and are graphically displayed using boxplots (Figure 6) and half-rose diagrams (Figure 7) for polyline lengths and orientations, respectively. It is apparent from Table 2 that the mean polyline lengths are significantly skewed by outlying data in both datasets. Therefore, the median is considered a more 16

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Figure 5: Final regional lineament maps for the (A) top-down and (B) bottom-up method17 as major and minor lineaments based on ologies. The bottom-up dataset is represented the metadata. Black lines represent the coast, rectangular inset area of HMNB Devonport c Crown Copyright and the eastern limit of the study area. This figure includes OS data and database right (2018).

ACCEPTED MANUSCRIPT

389 390 391 392 393 394

Range Skew 8700.06 4.09 10613.90 4.09 7706.14 5.82 8632.71 5.17

Kurtosis 27.24 27.45 58.47 44.12

SC

SD Median 491.83 191.05 678.27 238.54 346.15 143.95 470.13 180.28

IQR 289.93 429.84 183.48 261.45

M AN U

TD16 TD8 BU16 BU8

Mean 360.58 483.09 248.34 327.87

RI PT

Table 2: Summary statistics for the polyline length of each datasets. The mean is not considered a representative centre value due to the high skew. The range, median and kurtosis describe the non-normal distribution whilst the IQR provides a better comparison of dispersion than standard deviation. SD = standard deviation, IQR = Interquartile Range. The datasets are denoted as TD (top-down) and BU (bottom-up) where the numbers 8 and 16 refer to the number of compass directions used in post-processing.

appropriate measure of average polyline length. Due to the positively skewed distribution, the interquartile range, demarcated by the box in Figure 6 is used to reflect the dispersion within each dataset. It is apparent that there is less dispersion in the bottom-up method compared to the top-down method indicating a more consistent polyline length. Nevertheless, the top-down method produces a longer median polyline length.

AC C

EP

TE D

395

Figure 6: Boxplots demonstrating the variations in polyline lengths for the top-down (TD) and bottom-up (BU) methods. The TD method demonstrates a broad spread of the interquartile range and superior polyline lengths, whereas the BU method produces shorter lengths and has a smaller interquartile range.

396 397

The distribution of orientation is presented as half-rose diagrams in Figure 7. It is apparent from the rose diagrams that both methods capture a similar 18

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

Figure 7: Half-rose plots representing the population of polyline angles for top-down (TD) and bottom-up (BU) datasets. The data are binned at 6◦ angles. Both methods capture the same global trends with dominant populations trending ESE and SE with subordinate S populations.

19

ACCEPTED MANUSCRIPT

403

4. Discussion

406 407 408 409 410 411 412

413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429

SC

405

Two stand-alone OBIA methods for integrated lineament detection have been presented. Both methods can create similar data products but include different metadata which are further enhanced to include lineament length and orientation during post-processing. The initial use of line extraction in the CNL demonstrated by Middleton et al. (2015) is effective, especially where the TDR transform is applied to airborne magnetic data. This study extends the use of the method to TDR-transformed Total Count channel from airborne radiometric data and elevation data from an airborne LiDAR DTM.

M AN U

404

4.1. Application of the Tilt Derivative Image analysis often requires an effective enhancement to extract detailed information from an input dataset. The TDR transform has previously been applied for edge detection where the zero-contour corresponds to the edge of magnetic bodies and where the signal does not overlap with nearby sources (Fairhead et al., 2004; White and Beamish, 2011).

TE D

401

The use of TDR minima for detecting weakness zones in the bedrock, such as shear zones or fault zones, was described for worming techniques by Airo and Lev¨aniemi (2012). The selection of using minima does not preclude the use of maxima as dictated by the geological environment where fault zones may have enhanced magnetic, radiogenic or hardness properties. The detection of geological lineaments from TDR minima of airborne magnetic data using OBIA methods was first applied by Middleton et al. (2015) and compared to a similar worming technique to that of Airo and Lev¨aniemi (2012). Middleton et al. (2015) also applied to the TDR transform to airborne LiDAR DTM data but did not incorporate it into the semi-automated analysis.

EP

400

AC C

399

RI PT

402

trend in the orientation of lineaments. The dominant trends range from E through to SE with subordinate NE and N-S trends. Between the two methods there is little difference in the overall distribution, although, the top-down method appears to emphasise SE-trending lineaments more than the bottom-up method.

398

430 431 432

Here, the TDR transform has been successfully applied to airborne magnetic and LiDAR DTM, in addition to the Total Count channel of airborne 20

ACCEPTED MANUSCRIPT

439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464

465 466 467 468

RI PT

438

4.2. Top-down vs. Bottom-up Segmentation Top-down OBIA methods for geological lineament detection are effective at capturing the regional lineament network. The method does require user knowledge to define appropriate thresholds for segmentation, although these can be readily determined through a heuristic approach. Extending objects is computationally intensive and adds significant processing time to the top-down method. Furthermore, assigning object-ends from super-objects in a sub-level results in two nodes for extension which can be problematic for multi-branched objects. Additional extension after integrating objects is more sensitive as it requires another object-end to exist within proximity to that being extended. There is also no control over the direction of growth. Finally, the output metadata provides information on the data source of the lineament, which can be useful for geological interpretation.

SC

437

M AN U

436

TE D

435

In contrast, the bottom-up OBIA methodology requires less user knowledge since the whole image is segmented and then merged objectively. The method takes a slightly different approach by integrating the different lineness rasters into a composite lineness raster prior to segmentation. It also uses a border assessment approach to merge major objects into more extensive structures where they are interspersed with minor objects and create more robust lineaments. Some information is lost though integrating the lineness rasters, however, this, coupled with the border assessment step, makes the bottom-up method much more computationally efficient compared to top-down methods. Although some detail is lost during integration, metadata can be produced by defining thresholds for major and minor lineament classes.

EP

434

radiometric data. It is necessary to calculate the vertical derivative for the TDR transform using convolution, rather than a Fourier transform, for the LiDAR DTM and Total Count data. The resulting composite lineament networks are the first to use the TDR transform in this manner, and to successfully integrate multi-sourced data into a semi-automated analysis for geological lineament detection.

AC C

433

4.3. Comparison with Existing Geological Mapping Previous regional maps are often compiled for specific projects and tailored to focus on specific faults. The most complete regional map can be compiled from the 1:50 000 DiGMap data from the British Geological Survey 21

ACCEPTED MANUSCRIPT

470 471

(Figure 8). This map incorporates fault data from thirty mapping districts in the region that was originally mapped at 1:10 000 scale in the field and later collated and reduced to 1:50 000.

RI PT

469

472

475 476 477 478 479 480 481 482

SC

474

The regional 1:50 000 fault map is markedly inconsistent. Some districts have a high density of faults (Chulmleigh), whereas others are nearly completely devoid of them (Bodmin and Ivybridge). The poor outcrop exposure inland invariably makes conventional geological mapping difficult. However, regional variability is also influenced by lithostratigraphy (recognition of fault separation and geomorphological features e.g. Chulmleigh has a very favourable lithostratigraphy) and changes in historical emphasis and mapping techniques (data for some districts, e.g. Bodmin and Ivybridge was acquired over a century ago). Fault in upland granite areas are generally under-represented.

M AN U

473

483

487 488 489

490 491 492 493 494 495 496 497 498 499 500 501 502 503 504

TE D

486

4.4. Assessment of the Final Lineament Networks The final lineament maps presented in Figure 5 demonstrate two standalone OBIA methods for detecting lineaments. Both methods provide complete and consistent lineament maps which are created semi-automatically with minimal prior knowledge by the user. By integrating data from multiple sources, a composite lineament network is created based on geophysical, geological and geomorphological properties.

EP

485

The lineament datasets created from OBIA methods (Figure 5) are more comprehensive and consistent than the existing regional fault map (Figure 8). Furthermore, lineaments generated capture similar, but more distinct, structural trends. OBIA lineaments are comparatively shorter in length, although they have not been subjected to cartographic drafting as is the case with the 1:50 000 data.

AC C

484

The results from the two OBIA methods produce similar overall lineament patterns despite their fundamental differences. The extracted lineament networks are therefore considered to be robust and an accurate representation of the regional geological structure. The extension of image objects in the top-down method was initially devised by Middleton et al. (2015). The process is computationally intensive

22

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

Figure 8: The BGS 1:50 000 linear data for faults are presented. All faults have been grouped into observed or inferred with the number of objects noted in the legend. It is apparent that much of the BGS linework is based on inferred faults which are a likely combination of both geological and geomorphological information resulting in longer interpreted lengths than observed faults. The map also illustrates the inconsistencies across different mapping districts such as Chulmleigh (C), Ivybridge (I) and Bodmin (B). Furthermore, an inset is included to illustrate the orientation of all mapped (observed and c Crown Copyright and database right inferred) faults. This figure includes OS data (2018).

23

ACCEPTED MANUSCRIPT

506 507 508 509

despite producing longer lineaments compared to the bottom-up methodology. Statistical analysis of the orientation of lineaments from both methods concur with the trends recorded from field studies by Alexander and Shail (1995, 1996); Leveridge et al. (2002); Leveridge and Hartley (2006); Hughes et al. (2009); Nixon et al. (2011).

RI PT

505

510

513 514 515 516 517 518 519 520

Previous attempts at using OBIA methods for lineament detection have only been applied to a single input dataset. These include application to SAR (Marpu et al., 2008), Landsat (Mavrantza and Argialas, 2008) and LiDAR DTM data (Rutzinger et al., 2007), which restricts the mapping to lineaments that are only present at surface. Conversely, Middleton et al. (2015) use airborne magnetic data to determine whether lineaments are related to structures present in the deeper crust. The OBIA methods developed here are the first to integrate three multi-sourced input datasets for both surficial and subsurface lineament detection to create a more comprehensive and composite lineament network.

SC

512

M AN U

511

521

525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540

TE D

524

EP

523

The two contrasting OBIA methods presented here have different applications. For regional datasets, both methods can create a semi-automated lineament map. However, the bottom-up method provides a computationally efficient lineament map which sacrifices polyline lengths and output metadata. The metadata from the top-down method provides useful information on whether a lineament has a geomorphological expression (LiDAR DTM), may pertain to a deeper structure (magnetic data) or have been leached during alteration (radiometric data). Where a lineament has been detected in all three datasets, it is considered that this is more certain and likely to significant. Irrespective of the OBIA method selected, the common pre- and post-processing steps are crucial to ensuring quality of the final lineament maps. Careful pre-processing can mitigate the effect of artefacts in the input data, whereas thorough post-processing will remove any spurious lineaments generated during detection.

AC C

522

There is, of course, scope to develop an algorithm that finds a better middle ground between the two OBIA methodologies. Revised methods could attempt to optimise lineament lengths for the bottom-up method or reduce computational time of the top-down method.

24

ACCEPTED MANUSCRIPT

542 543 544 545 546

5. Conclusions Two stand-alone OBIA methods for the detection of geological lineaments from airborne geophysics and remote sensing datasets, to create a composite lineament network, are presented. OBIA methods for lineament detection are highly effective and can make use of top-down and bottom-up segmentation techniques.

RI PT

541

547

550 551 552 553 554

SC

549

The use of the TDR transform, and the assumption that lineaments are represented by minima in the magnetic, radiometric and LiDAR DTM data, is valid for the detection of geological lineaments in SW England, but does not preclude the algorithm from detecting maxima where appropriate. By applying the TDR transform to radiometric Total Count and LiDAR DTM data using convolution to calculate the vertical derivative, comparable datasets can be produced with similar properties for lineament detection.

M AN U

548

555

559 560 561 562 563 564 565 566 567 568 569 570

571

572 573 574 575

TE D

558

By integrating data from multiple sources, a comprehensive and composite lineament network is created based on geophysical, geological and geomorphological properties. The top-down method provides metadata that can aid geological interpretation and help determine the certainty and significance of a lineament. The bottom-up method is more computationally efficient but produces shorter lineament lengths and less detailed metadata.

EP

557

The further development of a top-down OBIA method for geological lineament detection allows the effective integration of multi-sourced datasets. The newly developed bottom-up OBIA method provides a more computationally efficient means of detecting lineaments across multiple datasets with less user input required at the expense of lineament length. Both OBIA methods produce similar composite lineament networks, which capture the orientations of existing mapped structures, and are therefore considered to be a robust representation of the regional structural geology.

AC C

556

Acknowledgements We would like to thank three anonymous reviewers for comments that have improved this manuscript. The work was completed as part of doctoral research by CMY funded by the British Geological Survey (BGS) and the GW4+ Natural Environment Research Council (NERC) Doctoral Training 25

ACCEPTED MANUSCRIPT

582

RI PT

581

SC

580

M AN U

579

TE D

578

EP

577

Partnership in collaboration with the Geological Survey of Finland (GTK). CMY is grateful to Eija Hyvnen (GTK) who advised on data preparation, and gave her time and computer facilities whilst CMY visited Rovaniemi, Finland. We are grateful to Jim White (BGS) for providing initial training in the Geosoft Oasis Montaj software and his constructive comments on an early draft of this manuscript. This paper is published with the permission of the Executive Director, British Geological Survey (NERC).

AC C

576

26

ACCEPTED MANUSCRIPT

Supplementary Information

583

586 587 588 589 590 591

The supplementary information herein relate to the article entitled “Two oppositely-directed Object-Based Image Analysis methods for semi-automated lineament detection” that were not included in the main article due to constraints on the article length. In Section 1 (S1) the software and inherent algorithms are stated and elaborated on where appropriate. S2 details the steps for pre-processing the airborne magnetic, radiometric and LiDAR data. S3 provides a description and discussion of how the polyline lengths and orientations were calculated.

RI PT

585

SC

584

592

594 595 596 597 598 599 600

S1 Software

M AN U

593

Pre-processing made use of a variety of software packages including Oasis Montaj (v8.5, Geosoft, Canada), the R project software (https://cran. r-project.org/) and ESRI ArcGIS software. Initial processing of the airborne magnetic and radiometric data was conducted in Oasis Montaj (v8.5, Geosoft, Canada). Furthermore, the MAGMAP extension and the raster package (v2.6–7, Hijmans 2017) were used within Oasis Montaj and R, respectively. Full details can be found in S2.

604 605 606 607 608 609 610 611 612 613 614 615 616 617

EP

603

All Object-based Image Analysis (OBIA) steps of the method are carried out in the eCognition software package, which uses the Cognitive Network Language (CNL) to create and manipulate image objects (v9.3, Trimble, Germany). The first step takes the input data individually and applies the line extraction algorithm to create a lineness raster. The similarity of pixels to a straight line are assessed within a rectangular window. The window is a kernel filter which is a combination of three stripes, a line stripe and two border stripes and defines lines based on an axial symmetry. The stripes are assessed for similarity whereby the line-to-border stripes and border-toborder stripes, are essential to the calculated lineness raster. A line is defined when the similarity of line-to-border stripes is below a given threshold and the border-to-border similarity is above a given threshold. In this study, the thresholds were kept to the default values. A lineness raster is calculated based on line quality. Line quality is defined in an 8-bit raster through Equation 2. The resultant raster has a range of 0–255 with higher values defined where a lineament is highly likely to exist.

AC C

602

TE D

601

618

27

ACCEPTED MANUSCRIPT

(sbb − bbt) (1 − slb) + 128 ∗ (1 − bbt) lbt

619 620 621

where LQ = Line Quality, ssb = similarity of border-border, bbt = borderborder threshold, slb = similarity of line-border, lbt = line-border threshold.

622

625 626 627 628 629 630 631

SC

624

The top-down method, uses the multi-threshold segmentation algorithm to create initial image objects, and chessboard segmentation algorithm to grow these objects prior to integration through merging and further growing using chessboard segmentation. The bottom-up method applies the multiresolution segmentation algorithm to a composite lineness raster before merging using the spectral difference algorithm. The resulting image objects are then refined using border assessment which is devised using object relationships within the CNL. The final step in eCognition is to convert image objects to skeletons and export a vector file.

M AN U

623

(2)

RI PT

LQ = 128 ∗

632

643

S2 Pre-processing

636 637 638 639 640 641

644 645 646 647 648 649 650 651

EP

635

AC C

634

TE D

642

The calculation of polyline length and orientation used a combination of ESRI ArcGIS software and R for the post-processing steps. Polyline lengths were calculated in the GIS. The orientation of each segment was calculated in the GIS before being opened in R and assigning a direction to each segment. The data were re-opened in the GIS and polylines were dissolved based on the common direction, preserving any metadata. The polyline length and orientation were then recalculated for the newly-dissolved polylines within the GIS. The quality assessment was conducted in a GIS to query the vector data for anomalously short polylines and visually examine the overall trends in the data.

633

The IGRF-corrected magnetic data were re-gridded from the original data using a minimum-curvature algorithm. The gridded data were transformed using the Reduction-to-Pole algorithm prior to applying the upward continuation algorithm to reduce cultural noise (Dentith & Mudge 2014). Upward continuation heights of 50, 100, 200, 300, 400, 500 and 750 m were tested. Although nearly all cultural noise sources were removed at 400 m, smallscale geological phenomena were smoothed. A continuation height of 50 m was taken forward as it was deemed that the linear noise sources became 28

ACCEPTED MANUSCRIPT

653 654 655 656 657

discontinuous at this height and small-scale geological phenomena could still be identified in the data. A 9 x 9 convolution filter was also applied to both the airborne magnetic and radiometric data to mitigate against minor N-S corrugations in the data caused by poor levelling between flight lines. These datasets were then transformed using the TDR filter using the MAGMAP extension.

RI PT

652

658

661 662 663 664 665 666 667 668 669

SC

660

The LiDAR data was pre-processed slightly differently. The data was first analysed in the R project software (https://cran.r-project.org) using the raster package (v2.6–7, Hijmans 2017). The resample algorithm coarsened the original 1 m gridded Digital Terrain Model (DTM) to 40 m pixels to match the resolution of the airborne magnetic and radiometric data. The data were then imported to Oasis Montaj as an ascii file and the 9 x 9 convolution filter was applied prior to transformation using the TDR transform. The TDR-transformed data were then opened in ESRI ArcGIS software using the Geosoft plug-in and converted to a standard file format that could be used across multiple software packages. Finally, the data were re-opened in R and resampled to the same extent.

M AN U

659

670

677

S3 Post-processing

674 675

678 679 680 681 682 683 684 685 686 687

EP

673

Due to the multi-branched nature of the final polylines, it was necessary to coerce these data into similar groups. The polylines were exploded at each vertex and the angle calculated for each resulting segment. The angles were then fitted to a range of 0 > a ≤ 180◦ . Directional groups were determined based on eight (N, NE, E, etc.) and sixteen (N, NNE, NE, etc.) compass directions. The groups were assigned based on the calculated angle and are centred to the compass directions. For the case of N-S orientations, the group includes data from 0
AC C

672

TE D

676

It should be noted that the 40 m pixel resolution is based on the original data collection of the airborne magnetic and radiomtric data. The 200 m spacing of N-S flight-lines resulted in 40 m pixels created by Beamish and White (2011). The LiDAR DTM was resampled to the same resolution for ease of comparison. There is obviously a loss of data where lineaments are smaller than the 40 m pixel resolution.

671

29

ACCEPTED MANUSCRIPT

689 690 691

the original metadata and the newly created directional groups. Finally, the angle of the dissolved polylines were calculated in addition to line length. Thus, four lineament datasets are represented in the statistical analysis for the two directional groups and the two methods.

RI PT

688

692

695 696 697 698 699

SC

694

The post-processing steps tested the use of sixteen- and eight-directional groups for comparing the length of polylines. It is an important consideration for linking exploded polyline segments into continuous polylines. The use of eight-directional groupings results in a longer median polyline length and is highest using the top-down OBIA method. Therefore, the top-down method with eight-directional post-processing is considered the best for creating long lineaments.

M AN U

693

700

704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724

TE D

703

It has been noted that there is a non-linear correlation between polyline length and orientation. In the sixteen-directional grouping, there are significant populations of polylines at ESE and SE orientations. However, for the eight-directional grouping, the ESE population is forced towards a more easterly trend. The eight-directional grouping produces longer polyline lengths. The reason for this is the probabilistic function that it requires fewer continuous pixels to create a long polyline at 0◦ , 45◦ , 90◦ , 135◦ than at other angles. It is therefore reasonable to determine that the population of ESE polylines are bending towards 90◦ due to increased polyline lengths when using an eight-directional grouping. Furthermore, bottom-up OBIA methods have a larger interquartile range in orientation compared to top-down methods. A larger interquartile range indicates that a broader spread of polyline orientations are captured. A broader range may suggest that the bottom-up method captures orientation in more detail, however, given the effect of bending this may be a function of the shorter polyline lengths being created.

EP

702

AC C

701

Final post-processing involved the quality assessment of the data and was undertaken in a GIS. Simply querying revealed anomalously short polylines existed that were <40 m. These were removed as it was deemed that a polyline shorter than the pixel width for the input data was likely to be an artefact and considered spurious. These may have occurred due to the removal of associated objects during the cleaning steps.

30

ACCEPTED MANUSCRIPT

731 732 733 734

735 736 737

738 739 740

741 742 743 744

745 746 747

748 749 750

751 752

753 754 755 756

RI PT

730

Airo, M.-L., Wennerstr¨om, M., jun 2010. Application of regional aeromagnetic data in targeting detailed fracture zones. Journal of Applied Geophysics 71 (2-3), 62–70. URL http://linkinghub.elsevier.com/retrieve/pii/ S0926985110000388

SC

729

M AN U

728

Airo, M.-L., Lev¨aniemi, H., 2012. Geophysical structures with gold potential in southern Finland. In: Gr¨onholm, S., K¨arkk¨ainen, N. (Eds.), Gold in Southern Finland: Results of GTK studies 1998-2011. Geological Survey of Finland, Special Paper 52. Geological Survey of Finland, pp. 227–244.

Alexander, A. C., Shail, R. K., 1995. Late Variscan Structures on the coast between Perranporth and St. Ives, Cornwall. Proceedings of the Ussher Society 8, 398–404. Alexander, A. C., Shail, R. K., 1996. Late- to post-Variscan structures on the coast between Penzance and Pentewan, South Cornwall. Proceedings of the Ussher Society 9, 72–78. Baatz, M., Hoffmann, C., Willhauck, G., 2008. Progressing from object-based to object-oriented image analysis. In: Blaschke, T., Lang, S., Hay, G. J. (Eds.), Object-Based Image Analysis: Spatial Concepts for KnowledgeDriven Remote Sensing Applications. Springer-Verlag, Berlin, pp. 29–42.

TE D

727

Beamish, D., White, J., 2014. TellusSW: airborne geophysical data and processing report. Tech. rep., British Geological Survey Open Report, OR/14/014.

EP

726

References

Beamish, D., White, J. C., 2011. Aeromagnetic data in the UK: A study of the information content of baseline and modern surveys across Anglesey, North Wales. Geophysical Journal International 184 (1), 171–190.

AC C

725

Blakely, R. J., Simpson, R. W., 1986. Approximating edges of source bodies from magnetic or gravity anomalies. Geophysics 51 (7), 1494. Blaschke, T., Burnett, C., Pekkarinen, A., 2004. Image Segmentation Methods for Object-based Analysis and Classification. In: de Jong, S. M., van der Meer, F. D. (Eds.), Remote Sensing Image Analysis: Including the Spatial Domain. Springer Netherlands, pp. 211–236. 31

ACCEPTED MANUSCRIPT

763 764

765 766 767 768

769 770 771

772 773 774

775 776 777

778 779 780 781

782 783

784 785

786 787

RI PT

762

SC

761

Debeglia, N., Martelet, G., Perrin, J., Truffert, C., Ledru, P., Tourlie, B., aug 2006. Semi-automated structural analysis of high resolution magnetic and gamma-ray spectrometry airborne surveys. Journal of Applied Geophysics 58 (1), 13–28. URL http://linkinghub.elsevier.com/retrieve/pii/ S0926985105000273

Diamant, E., 2004. Top-down unsupervised image segmentation (it sounds like an oxymoron, but it actually isn’t). In: Proceedings of the 3rd Pattern Recognition in Remote Sensing Workshop (PRRS’04), August 2004, Kingston University, UK.

M AN U

760

Dragut, L., Tiede, D., Levick, S. R., 2010. ESP: A tool to estimate scale parameter for multiresolution image segmentation of remotely sensed data. International Journal of Geographical Information Science 24 (6), 859–871. Eisank, C., Smith, M., Hillier, J., 2014. Assessment of multiresolution segmentation for delimiting drumlins in digital elevation models. Geomorphology 214, 452–464.

TE D

759

Fairhead, J. D., M. Green, C., Verduzco, B., Mackenzie, C., 2004. A new set of magnetic field derivatives for mapping mineral prospects. In: ASEG Extended Abstracts. No. 1. p. 1. Ferraccioli, F., Gerard, F., Robinson, C., Jordan, T., Biszczuk, M., Ireland, L., Beasley, M., Vidamour, A., Barker, A., Arnold, R., Dinn, M., Fox, A., Howard, A., 2014. LiDAR based Digital Terrain Model (DTM) data for South West England.

EP

758

Dearman, W. R., 1963. Wrench-faulting in Cornwall and south Devon. Proceedings of the Geologists’ Association 74 (3), 265–287.

AC C

757

Gayer, R. A., Cornford, C., 1992. The Portledge-Peppercombe Permian outlier. Proceedings of the Ussher Society 8, 15–18. Gerard, F., 2014. LIDAR QUALITY CONTROL REPORT Project: PM 1478. Tech. rep., Geomatics (Environment Agency, National Operations). Grebby, S., Cunningham, D., Naden, J., Tansey, K., 2012. Application of airborne LiDAR data and airborne multispectral imagery to structural

32

ACCEPTED MANUSCRIPT

794 795 796 797

798 799 800

801 802 803 804

805 806 807 808 809

810 811 812

813 814

815 816 817 818 819 820

RI PT

793

Hijmans, R. J., van Etten, J., Cheng, J., Mattiuzzi, M., Sumner, M., Greenberg, J. A., Perpinan Lamigueiro, O., Bevan, A., Racine, E. B., Shortridge, A., Ghosh, A., 2017. Raster: geographic analysis and modeling with raster data. R package, version 2.6-7.

SC

792

M AN U

791

Grebby, S., Field, E., Tansey, K., oct 2016. Evaluating the Use of an ObjectBased Approach to Lithological Mapping in Vegetated Terrain. Remote Sensing 8 (10), 843. URL http://www.mdpi.com/2072-4292/8/10/843

Holloway, S., Chadwick, R. A., 1986. The Sticklepath-Lustleigh fault zone: Tertiary sinistral reactivation of a Variscan dextral strike-slip fault. Journal of the Geological Society, London 143, 447–452. Hughes, S. P., Stickland, R. J., Shail, R. K., LeBoutillier, N. G., Alexander, A. C., Thomas, M., 2009. The chronology and kinematics of Late Palaeozoic deformation in the NW contact metamoprhic aureole of the Land’s End Granite. Geoscience in South-West England 12, 140–152. James, J. M., Moore, J. M., 1985. Multi-seasonal imagery studies for geological mapping and prospecting in cultivated terrain of S.W. England. In: Fourth Thematic Conference: ”Remote Sensing for Exploration Geology”, San Francisco, California, April 1-4, 1985. April 1 - 4, 1985. San Francisco, California, pp. 475–484.

TE D

790

Koike, K., Nagano, S., Ohmi, M., 1995. Lineament analysis of satellite images using a Segment Tracing Algorithm (STA). Computers & Geosciences 21 (9), 1091–1104.

EP

789

mapping of the upper section of the Troodos ophiolite, Cyprus. International Journal of Earth Sciences 101, 1645–1660.

AC C

788

Kresic, N., 1995. Remote sensing of tectonic fabric controlling groundwater flow in Dinaric karst. Remote Sensing of Environment 53 (2), 85–90. Lahti, I., Nyk¨anen, V., Niiranen, T., 2014. Gravity worms in the exploration of epigenetic gold deposits: new insights into the propectivity of the Central Lapland Greenstone Belt, Northern Finland. In: Niiranen, T., Lahti, I., Nyk¨anen, V., Karinen, T. (Eds.), Central Lapland Greenstone Belt 3D Modelling Project Final Report. Geological Survey of Finland, Ch. Chapter 1, pp. 8–17. 33

ACCEPTED MANUSCRIPT

827 828

829 830 831 832 833

834 835 836 837

838 839 840 841

842 843 844 845

846 847 848 849

850 851 852 853

RI PT

826

Lee, M., Morris, W., Harris, J., Leblanc, G., 2012. An automatic networkextraction algorithm applied to magnetic survey data for the identification and extraction of geologic lineaments. The Leading Edge (January), 26–31.

SC

825

Leveridge, B. E., Hartley, A. J., 2006. The Varisan Orogeny: the development and deformation of Devonian/Carboniferous basins in SW England and South Wales. In: Brenchley, P. J., Rawson, P. F. (Eds.), The Geology of England and Wales, 2nd Edition. The Geological Society, London, Ch. 10, pp. 225–256.

M AN U

824

Leveridge, B. E., Holder, M. T., Goode, A. J. J., Scrivener, R. C., Jones, N. S., Merriman, R. J., 2002. Geology of the Plymouth and south-east Cornwall area. Memoir of the British Geological Survey, Sheet 348. England and Wales. Mallast, U., Gloaguen, R., Geyer, S., R¨odiger, T., Siebert, C., 2011. Derivation of groundwater flow-paths based on semi-automatic extraction of lineaments from remote sensing data. Hydrology and Earth System Sciences 15 (8), 2665–2678.

TE D

823

Marpu, P. R., Niemeyer, I., Nussbaum, S., Gloaguen, R., 2008. A procedure for automatic object-based classification. In: Blaschke, T., Lang, S., Hay, G. J. (Eds.), Object-Based Image Analysis: Spatial Concepts for Knowledge-Driven Remote Sensing Applications. pp. 169–184.

EP

822

Lang, S., 2008. Object-based image analysis for remote sensing application: modeling reality - dealing with complexity. In: Blaschke, T., Lang, S., Hay, G. J. (Eds.), Object-Based Image Analysis: Spatial Concepts for Knowledge-Driven Remote Sensing Applications. Springer-Verlag, Berlin, pp. 3–27.

AC C

821

Masoud, A., Koike, K., 2006. Tectonic architecture through Landsat-7 ETM+/SRTM DEM-derived lineaments and relationship to the hydrogeologic setting in Siwa region, NW Egypt. Journal of African Earth Sciences 45 (4-5), 467–477. Masoud, A., Koike, K., 2017. Applicability of computer-aided comprehensive tool (LINDA: LINeament Detection and Analysis) and shaded digital elevation model for characterizing and interpreting morphotectonic features from lineaments. Computers & Geosciences 106, 89–100. 34

ACCEPTED MANUSCRIPT

860 861 862

863 864

865 866 867

868 869 870 871 872 873

874 875 876 877 878

879 880 881

882 883 884

885 886

RI PT

859

Middleton, M., Schnur, T., Sorjonen-ward, P., Hyv¨onen, E., 2015. Geological lineament interpretation using the Object-Based Image Analysis Approach: results of semi-automated analyses versus visual interpretation. Geological Survey of Finland, Special Paper 57, 135–154.

SC

858

Miller, H. G., Singh, V., 1994. Potential field tilt-a new concept for location of potential field sources. Journal of Applied Geophysics 32 (2-3), 213–217.

M AN U

857

Milligan, P. R., Gunn, P. J., 1997. Enhancement and presentation of airborne geophysical data. AGSO Journal of Australian Geology and Geophysics 17 (2), 63–75. Moore, J. M., Camm, S., 1982. Interactive enhancement of Landsat Imagery for structural mapping in tin-tungsten prospecting: a case history of the S.W. England Orefield (U.K.). In: International Symposium on Remote Sensing of Environment, Second Thematic Conference, Remote Sensing for Exploration Geology. Fort Worth, Texas, December 6 - 10, 1982, pp. 727–740.

TE D

856

Ni, C., Zhang, S., Liu, C., Yan, Y., Li, Y., 2016. Lineament Length and Density Analyses Based on the Segment Tracing Algorithm: A Case Study of the Gaosong Field in Gejiu Tin Mine, China. Mathematical Problems in Engineering 2016, 1–7. URL https://www.hindawi.com/journals/mpe/2016/5392453/

EP

855

Mavrantza, O. D., Argialas, D. P., 2008. Object-oriented image analysis for the identification of geologic lineaments. In: Blaschke, T., Lang, S., Hay, G. J. (Eds.), Object-Based Image Analysis: Spatial Concepts for Knowledge-Driven Remote Sensing Applications. Vol. 36. Springer-Verlag, Berlin, pp. 383–398.

Nixon, C. W., Sanderson, D. J., Bull, J. M., may 2011. Deformation within a strike-slip fault network at Westward Ho!, Devon U.K.: Domino vs conjugate faulting. Journal of Structural Geology 33 (5), 833–843.

AC C

854

O’Leary, D., Friedman, J., Pohn, H., 1976. Lineament, linear, lineation: Some proposed new standards for old terms. Geological Society Of America Bulletin 87 (10), 1463–1469. Paananen, M., 2013. Completed Lineament Interpretation of the Olkiluoto Region. Tech. Rep. October, Geological Survey of Finland. 35

ACCEPTED MANUSCRIPT

893 894 895

896 897 898 899

900 901 902 903

904 905 906

907 908 909 910

911 912 913

914 915 916

917 918

RI PT

892

Rogers, J. D., 1997. The interpretation and characterisation of lineaments identified from Landsat TM imagery of SW England. Phd thesis, University of Plymouth.

SC

891

Roberts, S., Andrews, J. R., Bull, J. M., Sanderson, D. J., 1993. Slowspreading ridge-axis tectonics: evidence from the Lizard complex, UK. Earth and Planetary Science Letters 116 (1-4), 101–112.

Rutzinger, M., Maukisch, M., Petrini-Monteferri, F., 2007. Development of Algorithms for the Extraction of Linear Patterns (Lineaments) from Airborne Laser Scanning Data. In: Proceedings of the Conference ’Geomorphology for the Future’, Obergurgl, 2007. pp. 1–8.

M AN U

890

Scheiber, T., Fredin, O., Viola, G., Jarna, A., Gasser, D., Lapi´ nska-Viola, R., 2015. Manual extraction of bedrock lineaments from high-resolution LiDAR data: methodological bias and human perception. Journal of the Geological Society of Sweden (GFF) 137 (4), 362–372.

TE D

889

Scrivener, R. C., Darbyshire, D. P. F., Shepherd, T. J., 1994. Timing and significance of crosscourse mineralization in SW England. Journal of the Geological Society, London 151 (4), 587–590. Shail, R. K., Alexander, A. C., feb 1997. Late Carboniferous to Triassic reactivation of Variscan basement in the western English Channel: evidence from onshore exposures in south Cornwall. Journal of the Geological Society, London 154 (1), 163–168.

EP

888

Pe˜ na, S. A., Abdelsalam, M. G., 2006. Orbital remote sensing for geological mapping in southern Tunisia: Implication for oil and gas exploration. Journal of African Earth Sciences 44 (2), 203–219.

Shail, R. K., Leveridge, B. E., 2009. The Rhenohercynian passive margin of SW England: Development, inversion and extensional reactivation. Comptes Rendus Geoscience 341, 140–155.

AC C

887

ˇ Silhav´ y, J., Min´ar, J., Mentl´ık, P., Sl´adek, J., 2016. A new artefacts resistant method for automatic lineament extraction using Multi-Hillshade Hierarchic Clustering (MHHC). Computers & Geosciences 92, 9–20. Smithurst, L. J. M., 1990. Structural remote sensing of south-west England. Proceedings of the Ussher Society 7, 236–241. 36

ACCEPTED MANUSCRIPT

925

926 927 928

929 930 931

RI PT

924

Verduzco, B., Fairhead, D., Green, C. M., MacKenzie, C., 2004. New insights into magnetic derivatives for structural mapping. The Meter Reader, 116– 119.

SC

923

Wang, J., Howarth, P. J., 1990. Use of the Hough transform in automated lineament detection. IEEE Transactions on Geoscience and Remote Sensing 28 (4), 561–567.

M AN U

922

White, J. C., Beamish, D., 2011. Magnetic structural information obtained from the HiRES airborne survey of the Isle of Wight. Proceedings of the Geologists’ Association 122 (5), 781–786.

TE D

921

EP

920

Sukumar, M., Venkatesan, N., Babu, C. N. K., 2014. A review of various lineament detection techniques for high resolution satellite images. International Journal of Advanced Research in Computer Science and Software Engineering 4 (3), 72–78.

AC C

919

37

ACCEPTED MANUSCRIPT Highlights

RI PT SC M AN U TE D EP

• •

Semi-automatic, Object-Based Image Analysis for geological lineament detection Two methods derived using oppositely directed segmentation techniques Integration of multi-sourced datasets creating a robust composite lineament network Uses the Tilt Derivative transform to standardise datasets prior to integration Combines airborne magnetic, radiometric and LiDAR datasets over SW England

AC C

• • •