A hierarchical approach for region-based image retrieval

We propose a hierarchical approach for region-based image retrieval, which is based on wavelet transform for its decomposition property similarity with human visual processing. First, automated image segmentation is performed fast in the low-low (LL) frequency subband of wavelet transform which shows desirable low resolution of image. In the proposed system, boundaries between segmented regions are deleted to improve the robustness of region-based image retrieval against uncertainty of segmentation. Second, region feature vector is hierarchically represented by information in all wavelet subbands and each feature component of a feature vector is a combined color-texture feature. Such feature vector captures the distinctive feature (e.g., semantic texture) inside one region finely. Through experiment results and comparison with other methods, the proposed method shows good tradeoff between retrieval effectiveness and efficiency as well as easy implementation for region-based image retrieval.