Tóm tắt: Many of the papers describing techniques and descriptors for content-based image retrieval (CBIR) describe their feature extractors (create image features) and query engine (calculate similarity) earlier. A large variety of features are applied to image retrieval. The interested question which features can be combined to obtain good results in a particular task. In this case, we propose a method of global features matching for image retrieval using three image features: color histogram, gabor wavelet and contourlet. The paper concludes with recommendation these three feature and the matching method of them. Finally, we show experiments demonstrating the good accuracy of this method.