In this work a novel technique for automatic edge enhancement and detection in synthetic aperture radar (SAR) images is proposed. The edge enhancement phase has been proven to be critical in heterogeneous SAR images, and the method proposed is a good solution that can be used to deal with this type of data. The method proposes a robust edge enhancement directly in the wavelet transformed domain based on a combination of wavelet coefficients at different scales. It does not require any type of pre filtering of input data, and is independent of the statistics of the input image. The adaptation capability of the method to very diverse scenarios with no need of a priori knowledge or settings is a useful feature in view of its integration in an unsupervised segmentation. General TermsImage Processing.

The enormous increase in the image database sizes, as well as its vast development in various applications, retrieval of images based on its content is an important research area with application to digital libraries and multimedia databases. Content Based Image Retrieval (CBIR) is one of the methods which use low level features like color, texture and shape information to retrieve the images. Effective texture feature is an essential component in any content based image retrieval systems in which Curvelet captures the texture features more efficiently than other spectral features like Gabor and wavelet. This paper describes an approach to achieve a rotation invariant CBIR by using Curvelet texture feature and make them into different segments of similar groups to improve the efficiency of retrieval. Combine all these features into a feature vector and retrieve the similar images based on similarity check. This approach gives better results even for the images found in different orientations.