A New Dynamic Texture Description Method With Local Binary Motion Patterns

With the rapid development of the computer technology, also the internet and multimedia technology, the processing and application of the video are more and more extensive. Dynamic Textures (DT) is a very important research work in the fields of pattern recognition and computer vision, and has an extensive application background in science research and engineering technology. DT is the image sequences with time-related stationary characteristics. They describe dynamic sceneries such as the dancing flag, joggling leaves, the waterfalls and the smokes. Feature extraction is one of the fundamental issues in pattern recognition. DT feature extraction refers to extract characteristic from image by some way that definite image processing techniques. A lot of feature descriptors have been suggested. However, most these DT feature descriptors are limited in simulations, and they have hardly been used in the real world applications. For instance, the VLBP is an advanced arithmetic of DT recognition, but it is difficult to extend VLBP to have a large number of neighboring points and this limits its applicability.The volume local binary pattern (VLBP) is an extension of the LBP operator widely used in ordinary texture analysis, combining the motion and appearance features. Derived from a general definition of texture in a local neighborhood, the local binary pattern (LBP) texture analysis operator is defined as a gray-scale invariant texture measure. It was first introduced as a complementary measure for local image contrast. The LBP method is an effective method in texture analysis and has many advantages. This article has carried on the outline of the LBP and the VLBP method's related concept and the theory.This paper proposes a new advanced method, which is based on the Local Binary Motion Patterns (LBMP) for extracting features of DT description and recognition by characterizing the motion and appearance features. The proposed method uses the Concatenated-LBMP histogram as the identifier of DT, which is concatenated by the LBMP histogram derived from motion features based on the Local Binary Pattern (LBP) block-matching criterion and the LBP histogram from appearance features. Experimental results based on the DynTex database show that the proposed method achieves a higher recognition performance than VLBP.