Analysis Of Texture Features And Applications In Image Processing

Texture as an important feature in image,including substantial visual inofrmation. The studys of texture features analysis as a hot and difficult spot is an importance approach to image perception,scenery recognize and target location. It also be widely used in the domain of content-based image retrieval,defect detection and medical image analysis.First of all,the paper introduces basic knowledge of texture features analysis,including some correlative home and abroad methods in existence.In-depth anatomy towards analytic methods of texture features analysis this paper makes some further explorations. The main innovations in the dissertation are:(1) To the unilateral and interference problems in the application of one single texture feature spectrum descriptor, this paper proposed a new feature spectrum descriptor.It will partition the image into texture area and smooth area by the texture spectrum.And choose the suitable features according to the characteristics of the image region itself. According to this character it can deftly describe the feature of the images which have dissimilar visual characteristics. In order to test the performance of the feature spectrum descriptor, this paper will apply it to the content-based image retrieval. Experiments show that the new feature spectrum descriptor has good retrieval efficiency, and a smaller time complexity.(2)This dissertation makes some study in the texture features of image and discuss the process of incomplete tree-structured wavelet transform. In this foundation, this paper studys the texture features extraction based on incomplete tree-structured wavelet transform. Some new types of texture features extraction is proposed in this paper. Eespecially. two methods of direction features extraction has been put forward. The paper also give the process and scope of the application. The experiments show the effectiveness of these methods.(3) To the singularity character of the textures and the edges of the image,this paper uses the Gaussian Mixture Model of contourlet coefficients to extract singularity signal,then apply the Otsu Automatic threshold method to detect the edge informations. The experiment shows improvements in performance for reducing the unimportant singular informations such as the small textures and the unattached points.