A texture image classification method based on multi-feature and random forest is proposed

aiming at the problem of poor recognition rate of distorted texture images by single texture feature and single classifier. The improved histogram of oriented gradient（HOG） feature extraction method and the gray level co-occurrence matrix of LBP image are used to extract features; the extracted feature matrix is cascaded into a new feature matrix

and the reduced-dimension fusion processing is carried out by using the principal component analysis method; feature matrix after reduced-dimension fusion is input into the random forest

and the final recognition rate is obtained by fusion voting. The experimental results on KTH-TIPS distorted texture image database show that the classification accuracy of distorted texture image is improved by using the method fusing multi-features and random forest

and it has better real-time performance.