Research On Content-Based SAR Image Retrieval Techniques

Synthetic Aperture Radar (SAR) is a kind of high-resolution imaging radar developed after World WarⅡ.With the development of SAR technology,how to retrieval the needed SAR images from the image database efficiently,rapidly and accurately becomes a hotspot in the field of SAR image application,so it is meaningful and applied to study content-based SAR image retrieval.After a comprehensive analysis on the key techniques of content-based image retrieval, this thesis mainly studies the applications of Nonsubsampled Contourlet Transform (NSCT) and semi-variances in content-based SAR image retrieval against the rotation problems,and the SAR image retrieval system was implemented based on Visual C++6.0 and Oracle 9i.A rotation-invariant algorithm is proposed based on NSCT for SAR image retrieval. By calculating the mean and standard deviation of image subbands decomposed by NSCT, the texture feature elements were extracted. For each scale, the feature elements were re-ordered in ascending by the sum of mean and standard deviation of each image subband, and the orientation sequences were adjusted by considering rotation-invariant to create orientation-weights and feature element-weights. Weighted Euclidean distance was used to improve retrieval efficiency. The experiment results demonstrate superiority of the proposed method. A rotation-invariant and flip-invariant algorithm is proposed based on semi-variances and Hu moments. A set of semi-variances with selected lags and directions represents the image spatial continuity information. In order to achieve rotation-invariant and flip-invariant, the semi-variances were re-ordered according to the spatial major direction determined by weighted semi-variances proposed in this paper. The re-ordered semi-variances combined with 12 moment invariants represent SAR image features. Canberra distance was used to measure the similarity of normalized image features. The experiment results demonstrate that the proposed method is outstanding and efficient.Finally, the SAR image retrieval system was implemented based on Visual C++6.0 and Oracle 9i using the above algorithms.This system was already deliveried to a research institute as a main part of"The SAR Image *** System".