Improved Fuzzy C Mean and Adaptive Weighted Majority Voting Based Land Cover Change Detection in Remote Sensing Images

Abstract

Change discovery along with distantly detected representation assumes a significant job in land spread mapping, procedure investigation and active data administrations. It is the way toward distinguishing the changed locations and changed sorts of Land Use and Land Cover (LULC) utilizing the distantly detected symbolism procured at various occasions. Various change identification procedures are advanced. The past work planned a Land Cover Change Detection (LCCD) scheme dependent on the mix of K-means algorithm and Adaptive Majority Voting (K means AMV) systems. But, it has an issue with False Alarm and furthermore to choose a group focuses in K-means calculation is troublesome. To handle of this issue the proposed framework structured an Improved Fuzzy C Mean (IFCM) clustering with Adaptive Weighted Majority Vote (AWMV) approach for Land Cover Change Detection (LCCD).In this scenario, Change Magnitude Image (CMI) from bitemporal pictures utilizing the picture distinction approach. At that point the pixel is utilized as the "focal pixel", and a versatile area is broadened continuously around the focal pixel utilizing two pre-characterized parameters (T1 and T2). The constraint tuning is implemented utilizing Simulated Annealing Inertia weight based Dragonfly Algorithm (SAIDA). Improved Fuzzy C Mean (IFCM) clustering is assumed for obtaining the mark of every pixel inside a versatile area. At long last, the mark of the focal pixel is refined through an Adaptive Weighted Majority Vote (AWMV) method. The whole CMI is handled and refined pixel-by-pixel, though the BCDM dependent on bi-transient remote detecting pictures are accessible. The outcomes demonstrate that the proposed accomplishes better execution contrasted and the current framework by means of accuracy, precision, recall and f-measure.

Keywords: Land cover,Change Magnitude Image (CMI), Adaptive Weighted Majority Vote (AWMV) and remote sensing images