Study Of Image Retrieval With Edge And Texture Feature

With the popularization of network and the development of multimedia technology, many images appear in the field of science and technology, people's daily work and life. How to retrieve the images from image database precisely and efficiently, which has been one important issue in the filed of image management and image application. Image retrieval has drawn the eyes of people.Content-Based Image Retrieval (CBIR) extracts visual features as retrieval features, such as image color, texture and shape, etc. Because of the lack of good directional selectivity and shift invariance, CBIR can't get good retrieval results. Using these features of DT-CWT, this paper resolved this problem and got the improvement of retrieval performance. And the method of extracting edge feature is sensitive to the illuminance, so it will limit the universal property of image database. This paper eliminates the drawback which coming from the change of illuminance by using phase congruency, in the result the retrieval system gets a improvement of robustness.After summarized image retrieval technology and presented rationale of double-tree complex wavelet transform (DT-CWT), this paper explains the designing method of the DT-CWT filters. The main contributions of this paper are summarized as follows:1. Designed the filters of DT-CWT. It not only keeps the multi-resolution and the analytic ability to time-frequency localizability, but also provides good directional selectivity, shift invariance and limited data redundancy which are lack in the traditional wavelet transform. It is just the shift invariance of DT-CWT solved the problem such as variation of energy in each level of coefficient for the shift of signal, which can not be solved using real wavelet transform.2. Extracted the texture feature of images using DT-CWT from six directions, then comparing with the traditional method - GLCM (Gray- Level Co-occurrence Matrix), to retrieval the sample image in the database, the improvement of retrieval performance shows the advantage of DT-CWT comparing with the GLCM. Six different distance metrics are tested for their performance of image retrieval to find the best one. 3. This paper presented a kind of edge feature extraction method. This method makes use of the concept which Fourier components have maximum congruency at feature points (phase congruency). Because of its dimensionless quality, it is invariant to changes in image brightness or contrast. Comparing with the traditional Harris method, experimental results prove that extracted features using phase congruency is invariant to image illuminance (or contrast).In the end, the work is summarized and the further research direction is pointed out.