Contourlet Features For 3D Surface Texture Fusion And Classification

As an important component of image understanding,analysis and recognition,texture analysis has a very wide range of applications in the field of pattern recognition and computer vision. Texture analysis composes of texture description,texture segment,texture classification,texture retrieval,among which texture classification and fusion are most component. In recent years, people have done many researches on 2D images applications using Wavelet and Beyond Wavelet, but Beyond Wavelet is not applied in 3D surface texture analysis. In this paper,contourlet-based texture feature extraction for purpose of classification and fusion are discussed and studied for 3D surface texture.In our paper, we first introduce 3D surface texture, contourlet features for 3D surface texture classification and fusion have been studied Systematically. Contourlet transform is a new kind of multi-scale and multi-resolution analysis tool and not only has characteristics of multi-resolution,locality and critical sampling which wavelet has,but also has the characteristics of multiple decomposition directions and anisotropy which wavelets lacking. So Contourlet transform has better sparse representation for image than wavelet transform. The basic theory and thought of Wavelet and Contourlet transform are introduced and we review the development of image fusion and classification techniques. We have three important tasks in our paper:Firstly, Contourlet transform is applied to fusion for 3D surface texture. With the contourlet-based features, we defined a fusion rule for generating new 3D surface texture images at arbitrary illumination directions based on the Lambertian model .Secondly, Contourlet transform is applied to achieve the feature vectors. With took the full advantage of directional information in different scales, texture features are extracted effectively. The classification test achieves good results.Thirdly, the fractal features combines with the Contourlet features and the result of the classification for 3D surface texture get more better.According to the experimental result, the proposed method can successfully classifies multiple different textures and distinguish the illumination directions. It is non-trivial to develop an application about 3D surface texture based on this the framework. Future work will focus on inventing new parameters to increase classification accuracy and more widely served for various kinds of textures.