ORIGINAL PAPER

Coupling of DEM and remote-sensing-based approachesfor semi-automated detection of regional geostructuralfeatures in Zagros mountain, Iran

Saied Pirasteh & Biswajeet Pradhan & Hojjat O. Safari &Mohammad Firuz Ramli

Received: 20 January 2011 /Accepted: 18 May 2011 /Published online: 31 May 2011# Saudi Society for Geosciences 2011

Abstract In recent years, remote-sensing data haveincreasingly been used for the interpretation of objectsand mapping in various applications of engineeringgeology. Digital elevation model (DEM) is very useful fordetection, delineation, and interpretation of geological andstructural features. The use of image elements for interpre-tation is a common method to extract structural features. Inthis paper, linear features were extracted from the LandsatETM satellite image and then DEM was used to enhancethose objects using digital-image-processing filteringtechniques. The extraction procedures of the linearobjects are performed in a semi-automated way. Photo-graphic elements and geotechnical elements are used asmain keys to extract the information from the satelliteimage data. This paper emphasizes on the application ofDEM and usage of various filtering techniques withdifferent convolution kernel size applied on the DEM.Additionally, this paper discusses about the usefulness ofDEM and satellite digital data for extraction of structuralfeatures in SW of Zagros mountain, Iran.

Keywords Remote sensing . Image processing . DEM .

Linear feature mapping . GIS . Zagros mountain

Introduction

Faults, folds, and lineaments interpreted from remotelysensed data are often used as indicator of major fractures innear surface (Ali Syed and Pirasteh 2003; Juhari andIbrahim 1997; Jana 2002). Many researchers have mappedand interpreted the structure of a region based on theinformation extracted from aerial photographs and satelliteimages (Berberian 1995; Sattarzadeh et al. 2000; Pradhan etal. 2006, 2010a, b, c; Pradhan 2010d; Ayazi et al. 2010;Farrokhnia et al. 2010; Pradhan and Youssef 2010; Safari etal. 2009; Youssef et al. 2009, 2011). The structural map canfurther be used for geoengineering applications such astunneling, dams, etc.

Ali and Pirasteh (2004) and Pirasteh and Syed Ahmad(2005) showed that using Laplacian convolution filter onband-4 of the Enhanced Thematic Mapper (ETM+) Landsatdata increase the ability to extract structural features.However, the potential of the DEM for the extraction ofthese feature and comparison with the satellite images havenot been performed earlier. Bolstad and Stowe (1994)produced DEMs from 1:40,000 scale aerial photographsand SPOT panchromatic pairs with automatic correlationtechnique and tested the accuracy of these DEMs and theirderivative surfaces (slope and aspect). Accuracies of DEMsderived from different data models were assessed by Li(1994). One of the DEMs was derived from photogram-metrically measured contour data, and the other was thegridded data from aerial photographs. As a result, hesuggested that contouring could be a better samplingstrategy for better accuracy.

S. Pirasteh :B. Pradhan (*)Institute of Advanced Technology,Spatial Numerical Modeling Laboratory,University Putra Malaysia,43400 UPM, Serdang, Malaysiae-mail: biswajeet24@gmail.come-mail: biswajeet@lycos.com

H. O. SafariFaculty of Earth Sciences, Golestan University,Sari, Iran

M. F. RamliFaculty of Environmental Studies,University Putra Malaysia,43400 UPM, Serdang, Malaysia

Arab J Geosci (2013) 6:91–99DOI 10.1007/s12517-011-0361-0

Generally, DEM uses for building 3D view of an area.DEM can be created from either stereo pairs derived fromsatellite data or aerial photographs or generated from digitalterrain elevation data. DEM can be readily combined withimage data (both optical and SAR) for a number ofdifferent purposes (Lillesand and Kiefer 1987). Forexample, DEM was used to calibrate and geocode SAR

data classification is very limited due to higher costinvolved in preparing the DEM.

Mostly, files in DEM format are taken from contourmaps. Various techniques of image preprocessing, integra-tion of ancillary data have shown positive results toimprove the filtering results. This study carried out toexplore the use of DEM in enhancing structural and

Fig. 1 Study area showing the Zagros mountains in southwest Iran

92 Arab J Geosci (2013) 6:91–99

geomorphology features with different filtering kernel sizeand classification techniques and further comparison withthe information extraction from the satellite image tounderstand the potential of the DEM for structural featuresextraction. This study demonstrates the comparison offiltering on DEM and ETM+ Landsat data.

Study area

The study area is a part of the Zagros mountains which liesin the southwest (Fig. 1) of Iran is one of the best exposedregions for geostructural features. This paper tries toexpress and answer the following question: (a) potentialof DEM for evaluation and extraction of geostructuralfeatures, (b) the usefulness of DEM with or without satelliteimage to extract the geostructural features with certaindegree of accuracy, (c) identifying of a suitable filtering andconvolution kernel to optimize the enhancement forextraction of the geostructural features from DEM, and (d)to generate a 3D image of the area in order to increase theaccuracy of the semi-automated interpretation in conjunctionwith several field data.

Data and methodology used

Digital image processing and DEM

Generally speaking, different methods may be applied onthe image to extract the geostructural and geomorphologyfeatures information (Lee and Pradhan 2006, 2007; Pirastehet al. 2008, 2009, 2010a, 2010b, 2011; Lee 2005; Giacintoet al. 2000; Pradhan et al. 2006; Pradhan and Lee 2009,2010a, b, c; Pradhan and Buchroithner 2010; Pradhan andPirasteh 2010). One of these methods is filtering andconvolution high-pass technique on the satellite image likeETM+ Landsat. However, this paper discusses aboutvarious convolution kernel size and matrix using differentfiltering and convolution techniques (i.e., high pass, lowpass, directional, and Laplacian) on DEM to compare themwith each other and further to ETM+. This will indicatehow DEM and which technique is more affected to extractstructural features.

Firstly, the contrast of six bands of the ETM+ data aredigitally enhanced. The images of all the bands arecompared with respect to contrast and definition ofstructural features. As a result of visual evaluation, LandsatETM+ band-4 data which record the information at thewavelength between 0.75 and 0.90 μm were selected forthis study, since it shows good contrast and displaystructural features like folds, faults, and lineaments ascompared with the other bands. However, the infrared

band-4 of ETM+ Landsat image has been considered forfiltering convolution kernel size too. Both directional andLaplacian convolution filter has been attempted withinENVI version 4.2 software environment. Figure 2 illustratesboth filters kernel size and marix applied on the ETM+satellite image.

The information extracted from both DEM and ETM+data are finally merged and compared in order tounderstand the potential of the DEM. Digital-image-processing filtering convolution techniques have beenapplied on the DEM, and different images are produced.The high-pass technique has been applied on DEM with 3×3, 5×5, 7×7, and 11×11 martix convolution kernel size(Fig. 3). The lineaments features are further enhanced.Low-pass filtering techniques have also been applied onDEM to enhance the objects (Fig. 4). Kernel size matrixapplied on DEM is 3×3, 5×5, and 7×7. Laplacian filteringconvolution kernel size of 3×3 and 5×5 have also beenattempted to extract the objects through a semi-automatedinterpretation on the DEM (Fig. 5). Finally, the directionaltechnique with zero angle with kernel size 3×3 and 5×5have been carried out to produce the images with betterenhancement and ability to interpret the objects (Fig. 6).

The source DEM has been produced using Rivertoolssoftware with text format extracted from aerial photographsand later generated digital topography map by IranianSurvey Organization (ISO). The data have been converted totext format using Microstation J software. Preprocessing,which is designed to remove any undesirable raster character-istics is produced by the initial processing of the raw data andremove noise and correct geometric distortions.

Fig. 2 Values produced by filtering techniques

Arab J Geosci (2013) 6:91–99 93

a 3*3 b 5*5

c 11*11

Fig. 3 Showing high-passtechniques applied on DEM

a 3*3 b 5*5Fig. 4 Showing low-passfiltering techniques on DEM

94 Arab J Geosci (2013) 6:91–99

The geometric correction of the ETM+ satellite imagewas performed by registering the image to 1:25,000 scaledigital topography map sheets by selecting 100 groundcontrol points. This has been carried out using ENVIsoftware. The resulted root mean square error was less than1 pixel (20 m). This was achieved by using the first-ordernearest neighborhood transformation method (Pradhan2010a, b, c, d, e, Pradhan 2011). This is because thedifference in contours lines from the digital topographymap provided by ISO is 20 m. A geometrically correctedraster image was then overlaid on the DEM to produce a3D image for increasing the accuracy and representing thelineaments and geomorphology features. The geostructuralfeatures were detected from the ETM+ data using imagetechniques elements after applying the digital-image-processing techniques. These techniques enhanced the

objects and further facilitate the analyst to make on-screendigitization. Furthermore, the objects have been mapped byinterpreting and using on-screen digitations along with thefield observations. This interpretation is defined as semi-automated interpretation. The keys for interpretation arephotography elements such as tone, shape, association, andgeotechnical elements such as erosion, topography, andvegetation. The same features are extracted from DEMseparately after applying all aforementioned filtering tech-niques. The linearity and sharp lines, slope, light gray-tonepixels, and elevation topography are main keys for interpre-tation on DEM. Subsequently, several field observations werecarried out in order to validate the extracted information fromETM+ and DEM data (Fig. 7). Finally, the extractedinformation were converted into the ARCGIS version9.2 environments to produce the final structural map.

a 3*3 b 11*11Fig. 5 Showing Laplacianfiltering techniques appliedon DEM

3*3 7*7Fig. 6 Showing directionalfiltering applied on DEM

Arab J Geosci (2013) 6:91–99 95

Results and discussion

In this paper, the most common way of presenting thegeostructural features and the potential of DEM using thedigital-image-processing techniques is disucssed. We applieddifferent filtering techniques and convolution kernel size onDEM. The geostructural features extracted from infraredband-4 of ETM+ Landsat enhanced the feature for structuralfeatures detection. This emphasizes and supports the infor-mation extracted from DEM. The information extracted fromDEM has shown a reasonable understanding and matcheswith the information extracted from ETM+.

To further enhance the geological structural information,filtering techniques such as Laplacian and directionalconvolution filters are applied to Band-4 Landsat ETM+data. The result shows that the number of filters of thevalues produced (Fig. 2) enhanced the general character-istics of folds, faults, lineaments, and thrusts of the area.The folds, faults, and lineaments in the study area are seenthrough different enhancements and filtering on images andsome are only seen in a directionally filtered images. Thereorganization of folds, faults, and lineaments on the imagesare based on vegetation linearity, tonal changes in the images,

drainage pattern, topographic breaks, landscapes, and discon-tinuity in the lithology, tectonic landforms such as klippe,nappe, fenster and scarps.

Different filtering kernel size on DEM reveals that thedirectional filter with 7×7 kernel is more powerful tointerpret the geostractural features visually and on-screendigital interpretation. The comparison of the differentgenerated images using various filtering techniques aredepicted in Table 1 to show the potential of the DEM forsemi-automated interpretation from DEM.

Moreover, the combinition of the extracted informationfrom DEM and ETM+ satellite data is very similar This wasfurther confirmed when the field observations informationhave been applied. In some pixels that do not allow theinterpretation of the structural features on ETM+, usingDEM can be a good combinition to extract more informationand validate the performances of ETM+. The DEM andETM+ were used to generate the 3D image of Zagrosmountain (Fig. 8) and overlaid the structural features on 3Dimage for increasing the accuracy and validating with theground observation.

When we applied high-pass filtering with convolutionkernel sizes 3×3 to 13×13 on DEM, it has been found thatthe structural features are more enhanced and can easilydetect the lineament features. But, when the same methodwas applied on ETM+ Landsat band 4, it did not suggest abetter enhancement. Although satellite data have its ownpotential to detect the objects and extract the information, ithas been shown that working on DEM using a high-passkernel size to extract structural features are more preferablethan the ETM+ data itself. It is because the ability of theanalyst to interpret the object increases mainly when thereis vegetation cover and makes it difficult to interpret thegeostructural and geomorphological features on the image.

The low-pass filtering process on DEM shows that whendifferent convolution kernel size was applid, no changes canbe seen on the original DEM. Furthermore, the Laplacianfiltering with different convolution kernel size onDEM showsthat if the kernel size is 3×3 and 5×5, the source DEM remaindark, but increasing the kernel size; the DEM starts to makethe features enhance. This enhancement is not as much aswhat we have seen in the high-pass convolution filter. Thedirectional filtering with zero angle for kernel size 3×3, 5×5,

Fig. 7 Showing lineaments and folds in Zagros mountains, southwestIran

Table 1 Comparison of the various filtering techniques applied on DEM

Convolution kernel size High pass Low pass Laplacian Directional

3×3 Very dark Normal as source DEM Very dark Very dark

5×5 Low enhanced Normal source as DEM Very dark Very low

7×7 Medium enhanced Normal source as DEM Low enhanced Low enhanced

9×9 High enhanced Normal source as DEM Medium enhanced Medium-high enhanced

96 Arab J Geosci (2013) 6:91–99

7×7, and 9×9 shows that the features would be enhancedwhen the kernel size increases and most of the features canclearly be seen on filtered DEM. But not as much as we couldsee in the high-pass filtered DEM.

Concluding remarks

Comparing the filtering techniques on DEM suggests thatthe high-pass filtering with 9×9 convolution kernel size is

Fig. 8 3D view and structural interpretation of Landsat ETM+ FCC 4-3-2 in Zagris Mountain, SW Iran. Approximate scale: 1:450,000. Fieldphotograph at right side of the figure illustrates slickensided surface and showing related movement along the fault plane

Fig. 9 Structural map of southwest Iran

Arab J Geosci (2013) 6:91–99 97

reliable to be used for the detection of geostructuralfeatures. Furthermore, the comparison between high-passfiltered DEM and ETM+ Landsat with band-4 infraredindicates that the combination of both raster data makes itpowerful and strengthens the ability to have a semi-automated interpretation. The information extracted fromthe ETM+, DEM, and field observations have been mergedtogether in GIS environment and produced a structural map(Fig. 9) for future geoengineering application purposes. Asa final conclusion, this study shows the advantages of theremotely sensed data and GIS techniques for detecting thelinear features by using semiautomatic filtering techniques.It also reveals that the DEM can be a very useful tool forthe study of geomorphology within the GIS environment.Similar studies can be adopted for different study areas withmountainous topography.

References

Ali Syed A, Pirasteh S (2003) Remote sensing and gis study oftectonics and net erosion rates in the Zagros Structural Belt,southwestern Iran. Mapp Sci Remote Sens 40(4):253–262

Ali Syed A, Pirasteh S (2004) Geological application of Landsat ETMfor mapping structural geology and interpretation: aided byremote sensing and gis. Int J Remote Sens 25(21):4715–4727

Ayazi MH, Pirasteh S, Arvin AKP, Pradhan B, Nikouravan B, MansorS (2010) Disasters and risk reduction in groundwater: Zagrosmountain southwest Iran using geoinformatics techniques. DisastAdv 3(1):51–57

Berberian M (1995) Master “blind” thrust faults hidden under theZagros Folds: active basement and surface morphotectonics.Tectonophysics 241:193–224

Bolstad PV, Stowe T (1994) An evaluation of DEM accuracy:elevation, slope and aspect. Photogramm Eng Remote Sensing60(11):1327–1332

Farrokhnia A, Pirasteh S, Pradhan B, Pourkermani M, Arian M (2010)A recent scenario of mass wasting and its impact on thetransportation in Alborz Mountains. Iran using geo-informationtechnology. Arabian J Geosci. doi:10.1007/s12517-010-0238-7

Giacinto G, Roli F, Bruzzone L (2000) Combination of neural andstatistical algorithms for supervised classification of remote-sensing images. Pattern Recogn Lett 21:385–397

Jana MM (2002) Application of remote sensing in the study ofgeomorphic processes and landform in piedmont zone ofDarjeeling Sub-Himalaya. Photonirvachak-J Ind 30(1–2):61–72

Juhari MA, Ibrahim A (1997) Geological application of LandsatThemetic Mapper imagery: mapping and analysing of lineamentsin NW Penisula Malaysia. Available at: www.gisdevelopment.net/aars/1997/. pp 1–7

Lee S (2005) Application and cross-validation of spatial logisticmultiple regression for landslide susceptibility analysis. Geosci J9(1):63–71

Lee S, Pradhan B (2006) Probabilistic landslide hazards and riskmapping on Penang Island, Malaysia. J Earth Syst Sci 115(6):661–672

Lee S, Pradhan B (2007) Landslide hazard mapping at Selangor,Malaysia using frequency ratio and logistic regression models.Landslides 4(1):33–41

Li Z (1994) A comparative study of the accuracy of digital terrainmodels (DTMs) based on various data models. ISPRS J Photo-gramm 49(1):2–11

Lillesand TM, Kiefer RW (1987) Remote sensing, and imageinterpretation, 2nd edn. Wiley, Singapore

Pirasteh S, Syed Ahmad A (2005) Lithostratigraphic study fromDezful to Brojerd-Dorood areas SW Iran using digital topography,remote sensing and GIS. Indian Petr Geol J 13(1):1–13

Pirasteh S, Woodbridge K, Rizvi SMA (2008) Geo-informationtechnology (GiT) and tectonic signatures: the River Karun &Dez, Zagros Orogen in south-west Iran. Int J Remote Sens30:389–395

Pirasteh S, Pradhan B, Mahmoodzadeh A (2009) Stability mappingand landslide recognition in Zagros mountain South West Iran: acase study. Disast Adv 2(1):47–53

Pirasteh S, Rizvi SMA, Ayazi MH, Mahmoodzadeh A (2010a) Usingmicrowave remote sensing for flood study in Bhuj Taluk, KuchchDistrict Gujarat, India. International Geoinformatics Researchand Development Journal 1(1):13–24

Pirasteh S, Safari HO, Pradhan B, Attarzadeh I (2010b) Litho-morphotectonics analysis using Landsat ETM data and GIStechniques: Zagros Fold Belt (ZFB), SW Iran. Intl Geoinf ResDev J 1(2):28–36

Pirasteh S, Pradhan B, Rizvi SM (2011) Tectonic process analysis inZagros mountain with the aid of drainage networks andtopography maps dated 1950–2001 in GIS. Arabian J Geosci 4(1–2):171–180. doi:10.1007/s12517-009-0100-y

Pradhan B (2010a) Remote sensing and GIS-based landslide hazardanalysis and cross-validation using multivariate logistic regressionmodel on three test areas in Malaysia. Adv Space Res 45(10):1244–1256

Pradhan B (2010b) Application of an advanced fuzzy logic model forlandslide susceptibility analysis. Intl J Comput Intel Syst 3(3):370–381. doi:10.2991/ijcis.2010.3.3.12

Pradhan B (2010c) Manifestation of an advanced fuzzy logic modelcoupled with geoinformation techniques for landslide suscep-tibility analysis. Environ Ecol Stat. doi:10.1007/s10651-010-0147-7

Pradhan B (2010d) Role of GIS in natural hazard detection, modelingand mitigation. Disast Adv 3(1):3–4

Pradhan B (2010e) Landslide susceptibility mapping of a catchmentarea using frequency ratio, fuzzy logic and multivariate logisticregression approaches. J Indian Soc Remote Sens 38(2):301–320

Pradhan B (2011) Use of GIS based fuzzy relations and its crossapplication to produce landslide susceptibility maps in three testareas in Malaysia. Environ Earth Sci 63(2):329–349

Pradhan B, Buchroithner MF (2010) Comparison and validation oflandslide susceptibility maps using an artificial neural networkmodel for three test areas in Malaysia. Environ Eng Geosci 16(2):107–126

Pradhan B, Lee S (2009) Landslide risk analysis using artificial neuralnetwork model focusing on different training sites. Intl J Phys Sci3(11):1–15

Pradhan B, Lee S (2010a) Delineation of landslide hazard areas onPenang Island, Malaysia, by using frequency ratio, logisticregression, and artificial neural network models. Environ EarthSci 60(5):1037–1054

Pradhan B, Lee S (2010b) Landslide susceptibility assessment andfactor effect analysis: back-propagation artificial neural networksand their comparison with frequency ratio and bivariate logisticregression modelling. Environ Modell Softw 25(6):747–759

Pradhan B, Lee S (2010c) Regional landslide susceptibility analysisusing back-propagation neural network model at CameronHighland, Malaysia. Landslides 7(1):13–30

98 Arab J Geosci (2013) 6:91–99

Pradhan B, Pirasteh P (2010) Comparison between predictioncapabilities of neural network and fuzzy logic techniques forlandslide susceptibility mapping. Disast Adv 3(2):26–34

Pradhan B, Youssef AM (2010) Manifestation of remote sensing dataand GIS on landslide hazard analysis using spatial-basedstatistical models. Arabian J Geosci 3(3):319–326. doi:10.1007/s12517-009-0089-2

Pradhan B, Singh RP, Buchroithner MF (2006) Estimation of stressand its use in evaluation of landslide prone regions using remotesensing data. Adv Space Res 37:698–709

Pradhan B, Lee S, Buchroithner MF (2010a) A GIS-based back-propagation neural network model and its cross application andvalidation for landslide susceptibility analyses. Comput EnvironUrban Syst 34:216–235

Pradhan B, Lee S, Buchroithner M (2010b) Remote sensing and GIS-based landslide susceptibility analysis and its cross-validation inthree test areas using a frequency ratio model. PhotogrammeFernerkund Geoinf 1:17–32. doi:10.1127/1432-8364/2010/0037

Pradhan B, Pirasteh S, Varatharajoo R (2010c) Enhancement ofautomated lineament extraction from IRS- 1B satellite imageryfor part of Himalayan region. Intl J Geo Inf 6(2):41–50

Safari HO, Pirasteh S, Pradhan B (2009) Upliftment estimation of theZagros Transverse Fault in Iran using geoinformatics technology.Remote Sens 1(4):1240–1256. doi:10.3390/rs1041240

Sattarzadeh Y, CosGrove JW, Vita-Finzi C (2000) The interplay offaulting and folding during the evoluation of the Zagros. In:Cosgrove, J.W.& Ameen. M.S. (eds) Forced Folds and Fractures.Special publication, 169, Geological Society, London. pp 187–196

Youssef AM, Pradhan B, Gaber AFD, Buchroithner MF (2009)Geomorphological hazard analysis along the Egyptian Red Seacoast between Safaga and Quseir. Nat Hazards Earth Syst Sci9:751–766

Youssef AM, Pradhan B, Sabtan AA et al (2011) Coupling of remotesensing data aided with field investigations for geologicalhazards assessment in Jazan area, Kingdom of Saudi Arabia.Environmental Earth Sciences, doi:10.1007/s12665-011-1071-3

Arab J Geosci (2013) 6:91–99 99