Differential Spatio-temporal Multiband Satellite Image Clustering using K-means Optimization With Reinforcement Programming

  • Irene Erlyn Wina Rachmawan Electronic Engineering Polytechnic Institute of Surabaya
  • Ali Ridho Barakbah Electronic Engineering Polytechnic Institute of Surabaya
  • Tri Harsono Electronic Engineering Polytechnic Institute of Surabaya


Deforestration is one of the crucial issues in Indonesia because now Indonesia has world's highest deforestation rate. In other hand, multispectral image delivers a great source of data for studying spatial and temporal changeability of the environmental such as deforestration area. This research present differential image processing methods for detecting nature change of deforestration. Our differential image processing algorithms extract and indicating area automatically. The feature of our proposed idea produce extracted information from multiband satellite image and calculate the area of deforestration by years with calculating data using temporal dataset. Yet, multiband satellite image consists of big data size that were difficult to be handled for segmentation. Commonly, K- Means clustering is considered to be a powerfull clustering algorithm because of its ability to clustering big data. However K-Means has sensitivity of its first generated centroids, which could lead into a bad performance. In this paper we propose a new approach to optimize K-Means clustering using Reinforcement Programming in order to clustering multispectral image. We build a new mechanism for generating initial centroids by implementing exploration and exploitation knowledge from Reinforcement Programming. This optimization will lead a better result for K-means data cluster. We select multispectral image from Landsat 7 in past ten years in Medawai, Borneo, Indonesia, and apply two segmentation areas consist of deforestration land and forest field. We made series of experiments and compared the experimental results of K-means using Reinforcement Programming as optimizing initiate centroid and normal K-means without optimization process.

Keywords: Deforestration, Multispectral images, landsat, automatic clustering, K-means.


Download data is not yet available.


James A. Shine and Daniel B. Carr, A Comparison of Classification Methods for Large Imagery Data Sets, JSM 2002 Statistics in an ERA of Technological Change-Statistical computing section, New York City, pp.3205-3207, 11-15 August 2002.

Bram Bakker & Jurgen Schmidhuber, Hierarchical Reinforcement Learning Based On Subgoal Discovery and Subpolicy Specialization. Technical report, IDSIA, 2003.

Ming tan, Multi-Agent Reinforcement Learning: independent vs cooperative agent, GTE Laboratories Incorporated.

C.Immaculate Mary, dr. S.V. Kasmir Raja, REFINEMENT OF CLUSTERS FROM K-MEANS WITH ANT COLONY OPTIMIZATION, Journal of Theoretical and Applied Information Technology, 2005–2009.

Abhijit Gosavi, A Reinforcement Learning Algorithm Based On Policy Iteration For Average Reward: Empirical Results With Yield Management And Convergence Analysis, Kluwer Academic Publishers, 2004.

Leslie Pack Kaelbling, Michael L. Littman, Andrew W. Moore. Reinforcement Learning: A Survey, Journal of Articial Intelligence, Artificial Research, 2004

D. Lu, Q. Weng, A survey of image classification methods and techniques for improving classification performance, International Journal of Remote Sensing, Vol. 28, No. 5, pp. 823-870, January 2007.

M. Govender, K. Chetty, V. Naiken and H. Bulcock, A comparison of satellite hyperspectral and multispectral remote sensing imagery for improved classification and mapping of vegetation, Water SA, Vol. 34, No. 2, April 2008.

Jasinski, M. F., Estimation of subpixel vegetation density of natural regions using satellite multispectral imagery, IEEE Transactions on Geoscience and Remote Sensing, Vol. 34, pp. 804–813, 1996.

James Theiler and Galen Gisler., A contiquity-enhanced k-means clustering algorithm for unsupervised multispectral image segmentation, Proc SPIE, 1997.

Kohei Arai, Ali Ridho Barakbah., Hierarchical K-means: an algorithm for centroids initialization for K-means, Reports of the Faculty of Science and Engineering, Saga University, Vol. 36, No.1, 2007.

Giri Tejaswi, Manual on Deforestation, Degradation, and Fragmentation using Remote Sensing and GIS, Forestry Department Food and Agriculture Organization of the United Nations, pp. 27-29, 2007.

Irene Erlyn Wina Rachmawan, Ali Ridho Barakbah, Ira Prasetyaningrum, Yuliana Setiowati, Reinforcement Programming: a function based Reinforcement Learning, The Third Indonesian- Japanese Conference on Knowledge Creation and Intelligent Computing (KCIC) 2014, Malang, Indonesia, March 25-26, 2014.

C.J. Veenman, M.J.T. Reinders, E. Backer, A Maximum Variance Cluster Algorithm, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 24, No. 9, pp. 1273-1280, 2002.

S. Ray, R.H. Turi, Determination of Number of Clusters in K-Means Clustering and Application in Colthe Image Segmentation, Proc. 4th ICAPRDT, pp.137-143, 1999.

Ali Ridho Barakbah, Kohei Arai, Identifying Moving Variance to make Automatic Clustering for normal dataset, Proc. IECI Japan Workshop 2004 (IJW 2004), Musashi Institute of Technology, Tokyo, 2004.

How to Cite
Rachmawan, I. E. W., Barakbah, A. R., & Harsono, T. (2015). Differential Spatio-temporal Multiband Satellite Image Clustering using K-means Optimization With Reinforcement Programming. EMITTER International Journal of Engineering Technology, 3(1), 133-152. https://doi.org/10.24003/emitter.v3i1.38