Multi-Dimensional Color Image Recognition and Mining Based on Feature Mining Algorithm

This paper introduced two algorithms of image feature mining: Histogram of Oriented Gradient (HOG) and gray-level co-occurrence matrix (GLCM). Then, the two algorithms were combined with support vector machine (SVM) model respectively to identify and classify color image. Simulation experiment was carried out in MATLAB R2018a software. The results showed that the feature texture image obtained by GLCM was more clear and accurate than that obtained by HOG; the GLCM based SVM had higher and more stable accuracy and shorter and more stable detection time.