Images research by content

This paper presents a new approach of Content Based Image Retrieval (CBIR), which combines the classification by SVM with the similarity measure using the GLCM, Gabor and Surf descriptors. The aim of this work has especially been concentrated on the level of improvement of these descriptors. So this improvement encourages us to propose a new approach of image research. The proposed approach is applied to synthetic and real data, and the results are compared to those obtained by other method. The compared results of those of algorithms of images search has shown the good performance of the proposed approach as well as the capacities of SVM as a classification tool of images.