Application Of Rice Paper Classification Study Based On Signal Processing And Statistical Methods In Image Texture Analysis

Chinese painting is the treasure of the Chinese nation. Following with the rapid development of China’s economy, the traditional Chinese painting fraud is rampant and inundant in recent years. Although the identification methods based on experts’ experience have been the main identification in the Chinese painting identification field, the experts’ identification has some disadvantages. Technology identification methods have been mentioned, but they are intrusive identifications which will cause certain damage to Chinese painting. Both the two identification methods already cannot meet the development needs of booming Chinese painting art market.Rice paper is the main carrier of traditional Chinese painting. In this paper we bring texture analysis which plays an important role in image processing into the study of rice paper’s classification. So we can provide the new non-intrusive and non-destructive technology identification for the Chinese painting identification. We mainly solve three problems in rice paper’s classification, namely the acquiring of texture images, the design of rice paper classification process and the implementation of rice paper texture analysis.Firstly, we design this special rice paper texture image acquisition system and provide the distinct texture images for rice paper’s classification. In our design the image nonuniformity has been increased to 0.58% by improving the light system. Comparing different rice paper fiber texture images we select the texture image under 54 magnifications as the final texture image. Secondly, after comparing the classification results of different types of texture analysis methods we choose LAWS texture and uncertainty texture spectrum to achieve the classification of rice paper. Comparing different kinds of classification algorithms, we find the SVM algorithm is the most suitable for rice paper classification. At the same time we find the optimal parameters of SVM algorithm. Finally, we choose LAWS texture and uncertainty texture spectrum as examples to describe how to use texture analysis method to achieve the classification of rice paper. The classification accuracy has reached to 75.7% and 78%. Based on the fact that the classification result is unsatisfactory when we use the two texture analysis methods separately, we improve the algorithm by combining the two methods closely so we can use both the advantages of them to improve the classification accuracy of rice paper. The classification accuracy has reached to 87.3% after improving the algorithm.The experimental results show that we cannot distinguish between different types of rice paper when we only use one single texture analysis method. The classification accuracy of rice paper has been distinctly improved after combining the two texture analysis methods. The accuracy of 87.3% has basically met the requirement for rice paper classification. In this paper, we extend the application area of the texture analysis and provide a new technical method for Chinese painting identification. The study of rice paper classification has positive significance for the scientific identification and economic value of Chinese painting.