Change Detection from Molti-Band Remote Sensing Image Based on Spectral Product Fusion and MRF Model

Abstract

Traditional remote sensing image change detection methods are difficult to detect complete change information by incomplete bands information, and change detection accuracy is not guaranteed. To solve this problem, the paper propose multi-band remote sensing image change detection method by spectral fusion and Markov Random Field. In the method, the difference image and ratio image are fused by spectral product fusion model, and get fusion image, all bands change information are used by iterative calculation, and obtain optimal change information. Experimental results show that the detection accuracy of the proposed method is superior to the current change detection methods, and good stability.