Rice Paper Classification Study Based on Signal Processing and Statistical Methods in Image Texture Analysis

Rice Paper Classification Study Based on Signal Processing and Statistical Methods in Image Texture Analysis

Haotian Zhai (Electronic Engineering Department, Jinan University, Guangzhou, China), Hongbin Huang (Electronic Engineering Department, Jinan University, Guangzhou, China), Shaoyan He (Electronic Engineering Department, Jinan University, Guangzhou, China) and Weiping Liu (Electronic Engineering Department, Jinan University, Guangzhou, China)
Copyright: © 2014 |Pages: 14
DOI: 10.4018/ijsi.2014070101
OnDemand PDF Download:
No Current Special Offers


Texture analysis plays an important role in image processing. In the field of texture analysis, the regular texture has been studied a lot, but the natural texture with complex backgrounds is less studied. This paper brings texture analysis into the study of rice paper's classification. First of all it shows the processing flow chart of rice paper classification. By comparing the different kinds of texture analysis methods it chooses the LAWS texture method and uncertainty texture spectrum method to achieve the rice paper classification. When it uses the two texture analysis methods separately, the classification accuracy of rice paper is lower, so it tries to combine the two texture analysis methods. The experimental results show that the classification result got with two combined texture analysis methods is better than that got with one single texture analysis method. The classification accuracy of rice paper has been distinctly improved after the combination of the two texture analysis methods.
Article Preview


Chinese painting is one of the oldest continuous artistic traditions in the world. It is the treasure of the Chinese nation and has won worldwide recognition. The component elements and composing style of Chinese painting have been sought after by people for hundreds of years. The main carrier of traditional Chinese paintings is the painting paper. The paper of different manufacturing processes and different time periods has different characteristics, so the texture of paper is also distinctive. Rice paper is the most commonly used traditional Chinese painting paper (Cao, 2012).The fiber of rice paper is rich, and the fiber texture mainly spreads on the surface of rice paper. Therefore, it is feasible to identify the traditional Chinese painting paper by extracting the fiber features of rice paper. At present, the judgment of rice paper is mainly based on experts’ identification (Yin, 2011), but this kind of identification is not perfect. Although technology identification methods have been mentioned, they are intrusive identifications which will cause certain damage to rice paper (Hu, Zeng, &Zhang, 2010). While the paper classification by means of fiber texture observation is a kind of nonintrusive classification method, it provides a feasible method for technology identification and has a positive scientific significance.

Texture is a kind of visual cues. It is also a kind of local pattern and arrangement rules repeated in the images. Texture is random, repetitive and regular. Traditional texture analysis methods can be divided into five categories, namely statistical method, geometric method, structure method, model method and signal processing method (Tuceryan &Jain, 1993).The statistical method is based on the gray of each pixel and its neighbors to study the statistical properties of the texture region. In 1973, Haralick (Haralick, Shanmugam, & Dinstein, 1973) firstly proposed a groundbreaking statistical method known as gray level co-occurrence matrix(GLCM).The geometric method uses geometric statistical characteristics to describe the texture. The application and development of this method are extremely limited, so the further research is less. The structure method analyzes the texture based on textons. The main structure texture analysis methods are the syntactic texture description method and the mathematical morphology method. The model method estimates the texture model by achieving the parameters of a texture image. The parameters of the model are always used for texture classification. In 1984,Pentland (Pentland, 1984) and his team did pioneering work on the model method. He pointed out that the fractal model was very suitable to describe the texture image. In addition, both the differential box proposed by Chaudhuri and Sarker (Sarkar, & Chaudhuri, 1994) and the extended fractal proposed by Kapan (Kapan, & Kuo, 1994) have a lot of applications in texture analysis. The signal processing method does some transformation based on time-frequency domain analysis and multi-scale transform. It classifies the texture with the transformation features. The most famous signal processing methods are the wavelet analysis introduced by Mallat (Mallat, 1989) in 1989 and the Gabor filter given by Dunnand Higgins (Dunn, Higgins, & Wakeley, 1994).

Although a lot of researches about texture analysis have been carried out, most of them care about the regular texture images. Complex texture images are relatively less studied (Liu, &Kuang, 2009). In addition, the current texture analysis focuses on the theoretical research. It pays little attention to the practical application. The study of texture analysis about specific rice paper has never been reported before. In this paper we compare a variety of texture analysis methods, then make a deep observation of rice paper fiber texture features, such as direction, thickness and other properties of texture. We take LAWS texture which belongs to the signal processing method and uncertainty texture spectrum which belongs to the statistical method as examples to describe how to use texture analysis method to achieve the classification of rice paper.

Complete Article List

Search this Journal:
Open Access Articles
Volume 10: 4 Issues (2022): 2 Released, 2 Forthcoming
Volume 9: 4 Issues (2021)
Volume 8: 4 Issues (2020)
Volume 7: 4 Issues (2019)
Volume 6: 4 Issues (2018)
Volume 5: 4 Issues (2017)
Volume 4: 4 Issues (2016)
Volume 3: 4 Issues (2015)
Volume 2: 4 Issues (2014)
Volume 1: 4 Issues (2013)
View Complete Journal Contents Listing