Colored texture classification with support vector machine and wavelet multiresolution analysis
Osama Hosam · 2015
Texture classification is essential part in automated industry and medical diagnosis. Traditional approaches for texture classification consider the gray scale image with intensity transition and variations in the texture image. Modern approaches use color information to add extra features to the classifier for stronger classification. Compared to Neural Networks, Support Vector Machines are more accurate and less computationally demanding technique for texture classification. In this paper we introduced texture image classifier based on wavelet transform and Support Vector Machine. Results showed high accuracy in classification when the color information is added compared to using grayscale images in texture classification.