International Journal of Computer
Trends and Technology

Research Article | Open Access | Download PDF

Volume 27 | Number 1 | Year 2015 | Article Id. IJCTT-V27P112 | DOI : https://doi.org/10.14445/22312803/IJCTT-V27P112

Classification of Hyper spectral Image Using Support Vector Machine and Marker-Controlled Watershed

Murinto, Nur Rochmah DPA

Citation :

Murinto, Nur Rochmah DPA, "Classification of Hyper spectral Image Using Support Vector Machine and Marker-Controlled Watershed," International Journal of Computer Trends and Technology (IJCTT), vol. 27, no. 1, pp. 70-75, 2015. Crossref, https://doi.org/10.14445/22312803/IJCTT-V27P112

Abstract

This research discuss the classification method observed which combined spatial information and spectral. There are three steps in the technique applied in this research. First, conduct the classification based on pixels hyperspectral image using suport vector machine (SVM). Second, the spacial contextual is used to rise the clasification result accuracy through the segmentation of hiperspektral image using the markercontrolled watershed method. Third, classsification based on pixel and image segmentation on the first step and the second, combined the result to aim the last map classification using the majority vote approach. The result finding obtained by using the hyperspectral image Aviris Indian Pines show the accuracy improvement compared with the classification using only the spectral information.

Keywords

Classification, hyperspectral image, marker-controlled watershed, support vector machine.

References

[1] R.Gaetano,”Hierarchical Models for Image Segmentation: From Color to Texture”, Tesi Di Dottorato, University Degli Studi Di Napoli, 2006.
[2] Shivani P.Deshmukh, Prof. Rahul D.Ghongade "Detection and Segmentation of Brain Tumor from MRI Image". International Journal of Computer Trends and Technology (IJCTT) V21(1):29-33, March 2015. ISSN:2231-2803.
[3] M.Fauvel, Y. Tarabalka, J.A. Benediktsson, Chanusso,”Advances in Spectral-Spatial Classification of Hyperspectral Images”, Publish in Proceeding IEEE 101, 3. 2013.
[4] D.A. Landgrebe, ”Hyperspectral image data analysis as a high dimensional signal processing problems”, IEEE Signal Processing Magazine, 19:17-28, 2002.
[5] G,Hughes,”On the Accuracy of Statistical Pattern Recognizers”, IEEE. Trans.Inf. Theory Vol.IT-14, no.1 pp.55 – 63, Jan. 1968.
[6] Tarabalka, Y., Benediktsson, J.A., Chanussot. Segmentation and Classification of Hyperspectral Images Using Watershed Transformation. 2010. Pattern Recognition 43, 7 (2010) 2367- 2379.
[7] P.Gamba,Hyperpsectral Dataset Available at http://dynamo.ecn.purdue.edu/biehhl/MultiSp
[8] C.Rodarmel,J.Shan, “Principal Component Analysis for Hyperspectral Image Classification”, Journal of the American Congress on Sureying and Mapping, 2, pp.115-122.2002.
[9] R.B. Cattel, ”The Scree Test of Number Factor”, Multivariat behavioral Research 1: 245-276, 1966.
[10] L. Vincent and P. Soille, “Watersheds in digital spaces: an efficient algorithm based on immersion simulations,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 13, no. 6, pp. 583–598, 1991
[11] Tarabalka, Y., Benediktsson, J.A., Chanussot, J., Tilton, J.C. Multiple Spectral-Spatial Classification Approach for Hyperspectral Data. 2010. IEEE Transaction on Geoscience and Remote Sensing, Vol.48, No.11, November 2010.Pattern Recognition 43, 7 (2010) 2367-2379.