基于Curvelet变换的图像检索算法研究-信号与信息处理专业论文.docxVIP

II II Abstract With the development of computer science, internet and multi-media technology, many great changes for human sociality, including economy, daily life, work, study etc., have taken place. With rapid improvement in the ability of image collecting, explosive growth in the amount of image storage, how to retrieve the useful image information quickly and exactly from vast images has become a research focus. Content-based image retrieval(CBIR) can make use of the low level visional features such as color, texture, shape and edge etc. It has been widely researched because of its advantages on overcoming many shortcomings of traditional retrieval appraoches. In CBIR, multi-resolution analysis can provide local time-frequency expression, and the possibility distribution models can fit the distribution of subbands in frequency fields, so it is the trends that make use of the combination of possibility distribution models and multi-resolution analysis. As a good multi-resolution analysis, wavelet was often used in texture image retrieval, but wavelet was found lack of directionality, curvelet was researched as an improvement because of the better directionality and more statistical information than wavelet. So far the features used on curvelet are low level statistical features which are simple, and the image texture information extracted are poor. For improving the retrieval ability we introduce the high level statistical features to CBIR. In this paper, we computed the generalized gaussian density and gray level-gradient co-occurrence matrix on the sub-bands of curvelet. Extensive experiments demonstrated the effectiveness of these two methods. The main works are as follows: Based on the multi-directionality of curvelet, we executed generalized gaussian density model on its sub-bands, and used the Newton-Raphson iteration to compute the parameters of probability density functions, then taken the Kullback-Leibler distance as the similarity measures on the parame