Performance evaluation of Artificial Neural Network, Support Vector Machine and Integrated Spectral Indices in satellite image classification

Articles in Press

Document Type : Research Paper

Authors

Department of Biosystems Engineering, Faculty of Agricultural Sciences, University of Guilan, Guilan, Iran

10.22124/cjes.2025.9183

Abstract

Satellite remote sensing is effectively used for environmental monitoring and change detection for the sustainable development of human society. While several methods exist for classifying satellite images, relatively few studies have focused on comparing these methods, especially considering the dimensional ratio and spatial distribution of the target phenomena. This study evaluates the performance of three classification methods including ANN, SVM, and an integrated approach that simultaneously uses three spectral indices of NDVI, GNDVI and, NDBI. The overall accuracy and kappa coefficient were calculated from the confusion matrix to statistically evaluate the three methods. Considering that statistical parameters are strongly sensitive to the dispersion and spatial distribution of the test points, a visual comparison was performed by overlaying the classified images with corresponding Google Earth imagery. Comparisons were made for several sample areas, which were categorized based on whether the land uses were integrated or scattered. Based on overall accuracy and kappa coefficient, the methods were ranked as SVM (97.36% and 0.9622), integrated spectral indices (94.06% and 0.9136), and ANN (93.42% and 0.9051). The visual comparison confirmed that SVM provided the best overall performance, consistent with the statistical results. Despite its lower overall accuracy, ANN was found more effective method in narrow areas compared to the other methods. Therefore, ANN is only recommended for detecting land uses with high levels of interference/integration with other features like rivers and roads that are surrounded by some other land uses.

Keywords

References

Adjorlolo, C, Mutanga, O, Cho, M & Ismail, R 2012, Challenges and opportunities in the use of remote sensing for C3 and C4 grass species discrimination and mapping. African Journal of Range & Forage Science, 29: 47-61.
Alipour, F, Aghkhani, M, Abasspour-Fard, M & Sepehr, A 2016, Demarcation and estimation of agricultural lands using etm+ imagery data (Case study: Astan Ghods Razavi Great Farm). Journal of Agricultural Machinery, 4: 244-254.
Bandyopadhyay, K, Pradhan, S, Sahoo, R, Singh, R, Gupta, V, Joshi, DK & Sutradhar, AK 2014, Characterization of water stress and prediction of yield of wheat using spectral indices under varied water and nitrogen management practices. Agricultural Water Management, 146: 115-123.
Bernstein, LS, Adler-Golden, SM, Sundberg, RL, Levine, RY, Perkins, TC, Berk, A, Ratkowski, AJ, Felde, G & Hoke, ML 2005, Validation of the QUick atmospheric correction (QUAC) algorithm for VNIR-SWIR multi-and hyperspectral imagery. In Algorithms and Technologies for Multispectral, Hyperspectral, and Ultraspectral Imagery XI, 5806: 668-678.
Buschmann, C & Nagel, E 1993, In vivo spectroscopy and internal optics of leaves as basis for remote sensing of vegetation. International Journal of Remote Sensing, 14: 711-722.
Caruana, R & Niculescu-Mizil, A 2006, An empirical comparison of supervised learning algorithms. In Proceedings of the 23rd international conference on Machine learning, pp. 161-168.
Dixon, B & Candade, N 2008, Multispectral landuse classification using neural networks and support vector machines: one or the other, or both? International Journal of Remote Sensing, 29: 1185-1206.
Ganasri, B & Ramesh, H 2016, Assessment of soil erosion by RUSLE model using remote sensing and GIS-A case study of Nethravathi Basin. Geoscience Frontiers, 7: 953-961.
Gandhi, GM, Parthiban, b, Thummalu, N & Christy, A 2015, NDVI: Vegetation change detection using remote sensing and gis–A case study of Vellore District. Procedia Computer Science, 57: 1199-1210.
Ghafari, S, Moradi, HR & Modarres, R 2018, Comparison of object-oriented and pixel-based classification methods for land use mapping (Case study: Isfahan-Borkhar, Najafabad and Chadegan plains). Journal of RS and GIS for Natural Resources, 9: 40-57.
Ghorbani, A, Aslami, F & Ahmadabadi, S 2016, Land and use mapping of Kaftareh of watershed of Ardabil using visual and digital processing of etm+ image. Natural Ecosystems of Iran, 6: 23-43.
Guerschman, JP, Hill, MJ, Renzullo, LJ, Barrett, DJ, Marks, AS & Botha, EJ 2009, Estimating fractional cover of photosynthetic vegetation, non-photosynthetic vegetation and bare soil in the Australian tropical savanna region upscaling the EO-1 Hyperion and MODIS sensors. Remote Sensing of Environment, 113: 928-945.
Haregeweyn, N, Berhe, A, Tsunekawa, A, Tsubo, M & Meshesha, DT 2012, Integrated watershed management as an effective approach to curb land degradation: a case study of the Enabered watershed in northern Ethiopia. Environmental Management, 50: 1219-1233.
Hazini, S & Hashim, M 2015, Comparative analysis of product-level fusion, support vector machine, and artificial neural network approaches for land cover mapping. Arabian Journal of Geosciences. 8: 9763-9773.
Huang, C, Davis, L & Townshend J 2002, An assessment of support vector machines for land cover classification. International Journal of Remote Sensing, 23: 725-749.
Jog, S & Dixit, M 2016, Supervised classification of satellite images. In 2016 Conference on Advances in Signal Processing (CASP) IEEE, 93-98.
Keno, K & Suryabhagavan, K 2014, Multi-temporal remote sensing of landscape dynamics and pattern change in Dire district, Southern Ethiopia. Journal of Geomatics, 8: 189-194.
Knight, JF, Lunetta, RS, Ediriwickrema, J & Khorram S 2006, Regional scale land cover characterization using MODIS-NDVI 250 m multi-temporal imagery: A phenology-based approach. GIScience & Remote Sensing, 43: 1-23.
Knorn, J, Rabe, A, Radeloff, VC, Kuemmerle, T, Kozak, J & Hostert, P 2009, Land cover mapping of large areas using chain classification of neighboring Landsat satellite images. Remote Sensing of Environment, 113: 957-964.
Kotsiantis, SB, Zaharakis, I & Pintelas P 2007, Supervised machine learning: A review of classification techniques. Emerging Artificial Intelligence Applications in Computer Engineering, 160: 3-24.
Li, C, Li, H, Li, J, Lei, Y, Li, C, Manevski, K & Shen, Y 2019, Using NDVI percentiles to monitor real-time crop growth. Computers and Electronics in Agriculture, 162: 357-363.
Lunetta, RS, Knight, JF, Ediriwickrema, J, Lyon, JG & Worthy, LD 2006, Land-cover change detection using multi-temporal MODIS NDVI data. Remote Sensing of Environment, 105: 142-154.
Madanian, M, Soffianian, AR, Koupai, SS, Pourmanafi, S & Momeni, M 2018, The study of thermal pattern changes using Landsat-derived land surface temperature in the central part of Isfahan Province, Iran. Sustainable Cities and Society, 39: 650-661.
Malik, S, Pal, SC, Das, B & Chakrabortty, R 2020, Intra-annual variations of vegetation status in a sub-tropical deciduous forest-dominated area using geospatial approach: A case study of Sali watershed, Bankura, West Bengal, India. Geology, Ecology, and Landscapes, 4: 257-268.
Markogianni, V, Dimitriou, E & Kalivas, D 2013, Land-use and vegetation change detection in Plastira artificial lake catchment (Greece) by using remote-sensing and GIS techniques. International Journal of Remote Sensing, 34: 1265-1281.
Maryantika, N & Lin, C 2017, Exploring changes of land use and mangrove distribution in the economic area of Sidoarjo District, East Java using multi-temporal Landsat images. Information Processing in Agriculture, 4: 321-332.
Mather, P & Tso, B 2016, Classification methods for remotely sensed data: CRC press, Florida, US, 357 p.
Mehta, A, Shukla, S & Rakholia, S 2021, Vegetation change analysis using normalized difference vegetation index and land surface temperature in Greater Gir Landscape. Journal of Scientific Research, 65: 1-6.
Mokhtari, MH & Najafi, A 2015, Comparison of support vector machine and neural networl classification methods in land use information extraction through Landsat TM data. Water and Soil Science, 19: 35-44.
Mountrakis, G, Im, J & Ogole, C 2011, Support vector machines in remote sensing: A review. ISPRS Journal of Photogrammetry and Remote Sensing, 66: 247-259.
Rahimi-Ajdadi, F 2022, Land suitability assessment for second cropping in terms of low temperature stresses using landsat TIRS sensor. Computers and Electronics in Agriculture, 200: 107205.
Rahimi-Ajdadi, F & Khani, M 2022. Multi-temporal detection of agricultural land losses using remote sensing and gis techniques, Shanderman, Iran. Acta Technologica Agriculturae, 25: 67-72.
Ramos, JF, Renza, D & Ballesteros, DM 2018, Evaluation of spectral similarity indices in unsupervised change detection approaches. Dyna, 85: 117-126.
Rawat, KS, Mishra, AK & Bhattacharyya, R 2016, Soil erosion risk assessment and spatial mapping using LANDSAT-7 ETM+, RUSLE, and GIS—A case study. Arabian Journal of Geosciences, 9: 1-22.
Richards, JA 1999, Remote Sensing Digital Image Analysis. Springer, Berlin, Germany, 494 p.
Rokni, K, Ahmad, A, Selamat, A & Hazini, S 2014, Water feature extraction and change detection using multitemporal Landsat imagery. Remote Sensing, 6: 4173-4189.
Sarp, G & Ozcelik, M 2017, Water body extraction and change detection using time series: A case study of Lake Burdur, Turkey. Journal of Taibah University for Science, 11: 381-391.
Sezgin, M & Sankur, B 2004. Survey over image thresholding techniques and quantitative performance evaluation. Journal of Electronic Imaging, 13: 146-165.
Sinha, P, Verma, NK & Ayele, E 2016, Urban built-up area extraction and change detection of Adama municipal area using time-series Landsat images. International Journal of Advanced Remote Sensing and GIS, 5: 1886-1895.
Slee, B 2007, Landscape goods and services related to forestry land use. Presented at Multifunctional Land Use: Meeting Future Demands for Landscape Goods and Services, Berlin, Heidelberg: Springer Berlin Heidelberg, pp. 65-82.
Tallón-Ballesteros, AJ & Riquelme, JC 2014, Data mining methods applied to a digital forensics task for supervised machine learning. In: Computational Intelligence in Digital Forensics: Forensic Investigation and Applications, Springer, pp. 413-428.
Talukdar, G, Sarma, AK & Bhattacharjya, RK 2020, Mapping agricultural activities and their temporal variations in the riverine ecosystem of the Brahmaputra River using geospatial techniques. Remote Sensing Applications: Society and Environment, 20: 100423.
Toosi, NB, Soffianian, AR, Fakheran, S, Pourmanafi,S, Ginzler, C & Waser, LT 2019, Comparing different classification algorithms for monitoring mangrove cover changes in southern Iran. Global Ecology and Conservation, 19: e00662.
USGS, 2018. Landsat missions. Landsat collection 1. https://www.usgs.gov/landsat-missions/landsat-collection-1#publications: Earth Resources Observation and Science (EROS) Center.
Varshney, A 2013, Improved NDBI differencing algorithm for built-up regions change detection from remote-sensing data: an automated approach. Remote Sensing Letters, 4: 504-512.
Vibhute, AD, Kale, K, Dhumal, RK & Mehrotra, S 2015, Hyperspectral imaging data atmospheric correction challenges and solutions using QUAC and FLAASH algorithms. International Conference on Man and Machine Interfacing (MAMI), 1-6.
Wu, D, Qu, JJ & Hao, X 2015, Agricultural drought monitoring using MODIS-based drought indices over the USA Corn Belt. International Journal of Remote Sensing, 36: 5403-5425.
Xu, H 2008, A new index for delineating built‐up land features in satellite imagery. International Journal of Remote Sensing, 29: 4269-4276.
Zambrano, F, Lillo-Saavedra, M, Verbist, K & Lagos, O 2016, Sixteen years of agricultural drought assessment of the BioBío region in Chile using a 250 m resolution Vegetation Condition Index (VCI). Remote Sensing, 8: 530.
Zhang, Z, Verbeke, L, De Clercq, E, Ou, X & De Wulf, R 2007, Vegetation change detection using artificial neural networks with ancillary data in Xishuangbanna, Yunnan Province, China. Chinese Science Bulletin, 52: 232-243.
Zhong, L, Hawkins, T, Biging, G & Gong, P 2011, A phenology-based approach to map crop types in the San Joaquin Valley, California. International Journal of Remote Sensing, 32: 7777-7804.
Caspian Journal of Environmental Sciences

Articles in Press, Corrected Proof
Available Online from 06 November 2025

Files

Share

How to cite

Statistics