A Comparative Study on Image Edge Enhancement for Synthetic Aperture Radar (SAR) Images

This paper is based on the glimpse of a comparative analysis of image edge enhancement techniques of synthetic aperture radar (SAR) images. The importance of edge enhancement is very much significant for SAR image recognition and other purposes. The first portion of this paper is based on a novel edge enhancement algorithm based on 2D wavelet transform. This method also signifies the motive to suppress the speckle noise from the image. Relying on the features of wavelet speckle image and log-compressed and transformed image, the result image is normalized, and finally the point wise maximum and multiplication has been taken to get the enhanced edge detection result. The second method is organized by taking sobel edge detection of SAR image. The resultant edge detected image is improved via a fuzzy logic based edge enhancement algorithm. Genetic algorithm based optimization is introduced in the next portion for further modification and enhancement of the edge detected and fuzzy-modified SAR image, which is finally re-enhanced by applying relaxed median filter on the image. Finally, the enhanced SAR edge image outputs for our two approaches are compared with different traditional edge detection techniques and the corresponding Root Mean Square Error (RMSE), normalized absolute error (NAE) and normalized cross-correlation error (NCC) have been computed and the comparative analysis is furnished. Keywordsedge enhancement, synthetic aperture radar (SAR) image, 2D wavelet transform, Fuzzy Logic based edge enhancement, Genetic algorithmic optimization in edge enhancement