Open AccessA Forest Change Detection using auto Regressive Model-Based Kernel Fuzzy Clustering
Author(s) -
Ms. Madhuri B. Mulik,
V. Jayashree,
Pandurangarao N. Kulkarni
Publication year - 2019
Publication title -
international journal of engineering and advanced technology
Language(s) - English
Resource type - Journals
ISSN - 2249-8958
DOI - 10.35940/ijeat.a1009.1291s619
Subject(s) - autoregressive model , fuzzy logic , change detection , kernel (algebra) , cluster analysis , pixel , artificial intelligence , computer science , pattern recognition (psychology) , fuzzy clustering , remote sensing , segmentation , image segmentation , land cover , data mining , geography , mathematics , statistics , land use , engineering , civil engineering , combinatorics
Satellite images are used for applications related to the forest change detection, forest cover management, and so on as remote sensing provides the rich source of information for change detection. In this paper, the vegetation indices play a major role in extracting the useful information from the satellite images and the commonly employed indices. This paper analyzes the imagery data from the remote sensing satellites for detecting the changes in the forest over the year’s 2007-2017 using the pixel-based Bhattacharya distance. The indices from the satellite images are fed to the automatic segmentation model using the proposed Kernel Fuzzy Auto regressive (KFAR) model, which is the modified Kernel Fuzzy C-Means (KFCM) Clustering algorithm with the Conditional Autoregressive Value at Risk (CAVIAR). The forest change detection using the pixel-based Bhattacharya distance follows the segmentation, and the experimentation reveals that the proposed method acquired the minimal MSE and maximal accuracy of 0.0581 and 0.9211.