A Fuzzy Scale-Space Approach to Feature-Based Image Representation and Retrieval

摘要

We propose an image indexing and retrieval method which is based on the multiscale image analysis theory in conjunction with fuzzy image feature extraction. The main idea is based on the assumption that the fundamental cues for image description such as shape and textures should be considered together within a unified model. Here the multiscale analysis is modeled by a differential morphological filter, and the feature are extracted by a multiscale fuzzy gradient operation applied to the detail images, which are the differences between images at successive scales. Experiments with large image databased and comparisons with classical methods are reported .