Research On Classification Of Texture Images

With technological development, the information content in multimedia database is more and more larger and the speed turns more and more quickly. It provides a lot of information also to us. However, it also brings a lot of troubles to us. The problem is how to sort and classify information in such a database.Multimedia database is so large that we retrieve data become more difficult. Rely on traditional methods is not meet users needs. In recent years, Content-based Image Retrieval (CBIR) has made great significant progress. It analyses the content of the image and gets characteristic. Then it uses image feature to build the indexing to retrieval image. That retrieves more effective.CBIR retrieves the content's color, texture, shape, and other visual features. Texture is the most notable feature of the visual feature. Texture features does not get the full use because of its difficulty. This article mainly research texture splitting and CBIR.Splitting texture is the key steps to analyses the image. We get the feature of the visual feature after slitting image. This article discusses the general approach to analyses the division of texture. This article discusses GLGM (the grey level grows matrix) and research on the effect of distance and grey level. It uses GLGM to get texture features and use K-means algorithm to slit texture. It achieved a satisfactory result in the experiment.