Comparison of Object-Based Classifiers and Traditional Pixel-Based Classification Techniques Using Landsat Imagery

Saviour Mantey, M. S. Aduah

Abstract


Object-Based Image Analysis (OBIA) is becoming dominant in remote sensing image classification. Many supervised classification approaches have been applied to objects rather than pixels, and studies have been conducted to evaluate the performance of such supervised classification techniques in OBIA. This study compares both the pixel-based and object-based techniques in classifying Landsat imagery. Pixel-based image classifiers such as the Maximum Likelihood Classifier and object-based image classifiers such as; Support Vector Machine (SVM), Random Forest (RF) and Decision Tree (DT) were compared using Landsat 8 imagery. The findings indicate that the SVM and RF methods obtained 94.86% overall accuracy with 0.9323 kappa, and 93.60 % overall accuracy with 0.9150 kappa, respectively as opposed to 92.73 % overall accuracy with 0.9077 kappa, for the pixel-based approach. From the results of this study, it was observed that the pixel-based image classification was constrained because the image pixels are not true geographical objects and the pixel topology is limited. It was also observed that the pixel-based classification largely neglects the spatial and photo-interpretive elements such as texture, context, and shape, which leads to the classifier resulting in lower classification accuracies. In contrast, the pixel-based method, OBIA works on (homogenous) image segmentation objects and can use more elements in the classification. This study therefore recommends that when classifying Landsat imagery for projects with higher accuracy requirements, the OBIA methods should be considered.


Keywords


Object-Based Image Analysis, SVM, RF DT, Pixel-Based Classification.

Full Text:

PDF

References


Alpaydin, E., (2014), "Kernel Machines, Introduc-tion to Machine Learning", Massachusetts Institute of Technology Press, pp. 349-385.

Anderson, J. R., (1971), “Land use classification schemes used in selected recent geographic applications of remote sensing”, Photogra-mmetric Engineering and Remote Sensing, Vol. 37, No. 4, pp. 379-387.

Anderson, J. R., Hardy, E. E., Roach, J. T., and Witmer, R. E., (1976), “A Land Use and Land Cover Classification System for Use with Remote Sensor Data (revised)”, U.S. Geological Survey, Professional Paper 964, 28 pp.

Anon., (2010a), eCognition® Developer 8.64.0 user guide, URL:http://www.definiens.com/ Acc-essed: March 11, 2020.

Anon. (2006), “Machine Learning”, https://en.wi-kipedia.org/wiki/Machine_learning#cite_note-bishop2006-2. Accessed: March 11, 2020.

Anon. (2014), “Image Classification Techniques in Remote Sensing”, https://gisgeography.com-/image-classification-techniques-remotesens-ing/. Accessed: January 15, 2020.

Anon. (2019). “The 10 Best Machine Learning Algorithms for Data Science Beginners”, https://www.dataquest.io/blog/top-10machine learning-algorithms-for-beginners/, Accessd: March 11, 2020.

Bakker, W. H., Feringa, W., Gieske, A. S. M., Gorte, B. G. H., Grabmier, K. A., Hecker, C. A., Horn, J. A., Huurneman, G. C., Janssen, L. L. F., and Kerle, N. (2004), “Principles of Remote Sensing”, Institure of GeoInformation Science and Earth Observation, 3rd edition, the Netherlands, pp. 193-203.

Bortolot, Z., (1996), “Image classification”, URL: http://www.geog.ubc.ca/course/geog516/notes/classification.html/, Accessed: March 7, 2020.

Breiman, L., and Cutler, A., (2007), “Random Forest-Classification Description”, URL: http://statwww.berkeley.edu/users/breiman/RandomForest/cc_home. Accessed: March 25, 2020.

Breiman, L., Friedman, J., Stone, C., and Olshen, R., (1984), “Classification and Regression Tree” , Belmont, California, USA: Champman and Hall/CRC, URL: http://www.amazon.ca/exec-/obidos/redirect?tag=citeulike0920&path=ASIN/0412048418. Accessed: April 6, 2020.

Camps-Valls, G. and Bruzzone, L. (2005), “Kernel-based methods for hyperspectral image classification”, Institute of Electrical and Electronics Engineers Transactions on Geoscience and Remote Sensing, Vol. 43, No. 6, pp. 1351–1362.

Camps-Valls, G., Gomez-Chova, L., Munoz-Marı, J., Rojo-Alvarez, J. L. and Martınez-Ramon, M. (2008), “Kernel-based framework for multi-temporal and multi-source remote sensing data classification and change detection”, Institute of Electrical and Electronics Engineers Transactions on Geoscience and Remote Sensing, Vol. 46, No. 6, pp. 1822–1835.

Clinton, N., Holt, A., Scarborough, J., Yan, L., and Gong, (2010), “Accuracy Assessment of Object-Based Image Segmentation Goodness”. Photogrammetric Engineering and Remote Sensing, Vol. 76, No. 3, pp. 289-299.

Congalton, R. G. (1991), “A review of assessing the accuracy of classifications of remotely sensed data”, Remote Sensing of the Environment, Vol. 37, pp. 35–46.

Dehvari, A. and Heck, R. J., (2007), “Comparison of Object-Based and Pixel Based Infrared Airborne Image Classification Methods Using DEM Thematic Layer”, Journal of Geography and Regional Planning, Vol. 2, No. 4, pp. 086-096.

Duro, C. D., Franklin, E. S., and Dube, G. M., (2011), “A Comparison of Pixel-Based and Object-Based Image Analysis with Selected Machine Learning Algorithms for Classifica-tion of Agricultural Landscape Using SPOT-5 HRG Imagery”, Remote Sensing of the En-vironment, Vol. 118, pp. 259-272.

Eisank, C., Smith, M. J. and Hillier, J. K., (2014), “Assessment of Multi-resolution Segmen-tation for Delimiting Drumlins in Digital Elevation Models”, Geomorphology, Vol. 214, pp. 452 - 464.

Everingham, M., Van Gool, L., Willians, C. K., Winn, J., and Zisserman, A., (2010), “The Pascal Visual Object Classes (VOC) Chall-enge”, International Journal of Computer Vision, Vol. 8, pp. 303-338.

Fumo D. (2017), https://towardsdatascience.com-/types-of-machine-learning-algorithms-you-should-know-953a08248861, Accessed: Aug-ust 11, 2020.

Hay, G. J., Blaschke, T., Marceau, D. J., and Bouchard, A., (2003), “A Comparison of Three Image Object Methods for the Multi-Scale Analysis of Landscape Structure”. ISPRS Journal of Photogrammetry and Remote Sensing, Vol. 57, pp. 327-345.

Hay, G. J., Castilla, G., (2006), “Object-Based Image Analysis: Strengths and Weaknesses, Opportunity and Threats (SWOT)”, 1st International Conference on Object-Based Image Analysis (OBIA), Salzburg, Austria. pp. 4-5.

Hehman, S. V., (1997), “Selecting and Interpreting Measures of Thematic Classification Accur-acy”, Remote Sensing of the Environment, Vol. 62, No. 1, pp. 77-89.

Hossain, M. D. and Chen, D. (2019), “Segmentation for Object-Based Image Analysis (OBIA): A Review of Algorithms and Challenges from Remote Sensing Perspective”, International Society for Photogrammetry and Remote Sensing Journal of Photogrammetry and Remote Sensing, Vol. 150, pp. 115-134.

Hutchinson, C. F., (1982), “Techniques for Combin-ing Landsat and Ancillary Data for Digital Classification improvement”, Photogramm-etric Engineering and Remote Sensing, Vol. 48, pp 123-130.

Jenssen, L. L., and Middelkoop, F. H., (1992), “Knowledge-base crop classification of La-dsat Thermatic Mapper image”, Internatio-nal Journal of Remote Sensing, 13, pp. 2827-2837.

Kuhn, M., (2011), “The Caret Package”, URL: http://cran.rproject.org/web/packages/caret/vignette/caretTrain.pdf. Accessed: April 10, 2020.

Li, M., Ma, L., Blaschke, T., Cheng, L. and Tiede, D., (2016), “A Systematic Comparison of Different Object-Based Classification Tech-niques Using High Spatial Resolution Ima-gery in Agricultural Environments”, Inter-national Journal of Applied Earth Obser-vation Geoinformatics, Vol. 49, pp. 87-98.

Lillesand, T. M., and Kiefer, R. W., (1994), “Remote Sensing and Image Interpretation”, 3rd edition, John Wiley and sons, New York, USA, pp. 524-531, 585-613.

Ma, L., Cheng, L., Li, M., Liu, Y., and Ma, X., (2015), “Training Set Size, Scale and Features in Geographic Object-Based Image Analysis of Very High Resolution Unmanned Aerial Vehicle Imagery”, International Society for Photogrammetry and Remote Sensing, Journal of Photogrammetry and Remote Sensing, Vol. 102, pp. 14-27.

Ming, D., Ci, T., Cai, H., Li, L., Qiao, C., and Du, J., (2012), “Semivariogram-Based Spatial Bandwidth Selection for Remote Sensing Image Segmentation with Mean-Shift Algorithm”, Institute of Electrical and Electronics Engineers Geoscience Remote Sensing Letter, Vol. 9, pp. 813-817.

Qian, Y., Zhou, W., Yan, J., Li, W. and Han, L. (2014), “Comparing Machine Learning Clas-sifiers for Object-Based Land Cover Class-ification Using Very High-Resolution Imag-ery”, Remote Sensing, Vol. 1, No.7, pp. 153–168.

Rodriguez-Galiano, V. F., Chica-Olmo, M., Abaca-Hernandez, F., Atkinson, P. M., and Jegana-than, C., (2012), “Random Forest Classifica-tion of Mediterranean Land Cover Using Mu-lti Seasonal Imagery and Multi-Seasonal Texture”, Remote Sensing of Environment, Vol. 121, pp. 93-107.

Stehman, S. V., and Czaplewski, R. L., (1998), “Design Analysis for Thematic Map Accura-cy Assessment: Fundamental Principles”, Re-mote Sensing of Environment, Vol. 64, No. 3, pp. 331-344.

Story, M. and Congalton, R. G. (1986), “Accuracy assessment: A user’s perspective”, Photogra-mmetric Engineering and Remote Sensing, Vol. 52, pp. 397–399.

Tilahun, A. and Teferie, B., (2015), “Accuracy Assessment of Land Use Land Cover Classi-fication Using Google Earth”, American Journal of Environmental Protection, Vol. 4, No. 4, pp. 193-198.

Zhang, C. and Xie, Z. (2013), “Object-based Vege-tation Mapping in the Kissimmee River Watershed Using Hyperspectral Map Data and Machine Learning Techniques” Remote Sensing of Environment, Vol. 2, No. 33, pp. 233–244.


Refbacks

  • There are currently no refbacks.