A New Algorithm for SAR Imagery Similarity Measure Based on Local Gradient Ratio Pattern
Xiang De-lian · Dianzi xuebao · 2014
In this paper,a new similarity measure for SARimagery based on local gradient ratio pattern is proposed.Firstly,we extract the gradient ratio pattern for each pixel.Secondly,the local gradient ratio pattern histogram(LGRPH)is computed.Finally,the similarity is obtained by utilizing K-Ldiscrepancy to measure the distance of LGRPH.Experimental results based on simulated and real SAR images demonstrate that the proposed similarity measure is robust to the speckle noise and local gradient variation in SAR images.