Content based texture image classification
Xiao-Qing Shang, Song Guo-xiang, Biao Hou · 2004
An image can be regarded as a combination of different texture areas. Texture is usually defined as a certain local feature, or it measures the relation between pixels in a local area. It is also an important feature in content based image retrieval. Based on moment feature, a new method for content based texture image classification is proposed using support vector machine and directional information of the image, which combines the characteristics of Brushlet and wavelet transform. Experiments show our method is practicable and effective.