Implementation and comparison of classifiers for different hyperspectral dataset based on machine learning algorithms


  • Kishore M
  • S. B. Kulkarni



In these paper better classification accuracy techniques is proposed for flower and land cover hyperspectral dataset. Initially flower dataset is considered in which newly proposed improved particle swarm optimization is implemented and compared with particle swarm optimization and K means algorithm followed by land cover dataset is considered in which proposed random forest algorithm is compared with support vector machine and k means and Navie Bayes classifiers. In both the hyperspectral dataset proposed methods gives good classification results in terms of accuracy.


[1] Vitousek, P.M., 1994. Beyond global warming: ecology and global change. Ecology 75, 1861-1876.

[2] Chandan, M. C, Vinay, S., Bharath H. Aithal, Ramachandra, T.V. 28-30th December 2016. Land use assessment and urban growth monitoring in Hyderabad region, India.

[3] Ramachandra, T. V., Aithal, B. H., Vinay, S., 2013. Land use land Cover dynamics in a rapidly urbanising landscape SCIT Journal, vol: 13: Pages 1-1.

[4] Sharifah Lailee Syed Abdullah, HamirulAiniHambali, Nursuriati Jamil, "Segmentation of Natural Images Using an Improved Thresholding-based Technique", International Symposium on Robotics and Intelligent Sensors 2012, Science Direct.

[5] S. Jenicka, A. Suruliandi, “A Textural Approach for Land Cover Classification of Remotely Sensed Image", International Journal of Computer Applications (0975 – 8887) Volume 92 – No.10, April 2014.

[6] Chandan, M. C, Vinay, S, Bharath H. Aithal, Ramachandra T.V, "Land use assessment and Urban Growth Monitoring in Hyderabad region, India", Conference on Conservation and Sustainable Management of Ecologically Sensitive Regions in Western Ghats, THE 10TH BIENNIAL LAKE CONFERENCE.

[7] V.F. Rodriguez-Galiano, B. Ghimire, J. Rogan, M. Chica-Olmo, J.P. Rigol-Sanchez, "An assessment of the effectiveness of a random forest classifier for land-cover classification", ISPRS Journal of Photogrammetry and Remote Sensing, 2012, Science Direct.

[8] Cavallaro, G.; Dalla Mura, M.; Benediktsson, J.A.; Bruzzone, L., "A comparison of self-dual attribute profiles based on different filter rules for classification", in Geoscience and Remote Sensing Symposium (IGARSS), 2014. IEEE International, pp.1265-1268, 2014.

[9] Rebetez, J.; Tuia, D.; Courty, N., "Network-Based Correlated Correspondence for Unsupervised Domain Adaptation of Hyperspectral Satellite Images", in Pattern Recognition (ICPR), 2014.22nd International Conference, pp.3921-3926, 2014.

[10] Jianjun Liu; Zebin Wu; Zhihui Wei; Liang Xiao; Le Sun, "Spatial-Spectral Kernel Sparse Representation for Hyperspectral Image Classification", in Selected Topics in Applied Earth Observations and Remote Sensing, IEEE Journal, vol.6, no.6, pp.2462-2471, 2013.

View Full Article: