Abstract: The workpiece identification method (WIM) based on improved SURF algorithm is proposed against the SURF which prone to match wrongly. The method first apply the Hessian matrix of SURF to detect feature points. Then, the Euclidean distance is utilized to match. Finally, the improved RANSAC algorithm is used to assume the exact feature point pair mathematical model, and find the "inner point" that conforms to the mathematical model, so as to improve the accuracy of feature matching. The purpose is to improve the efficiency of workpiece recognition in the industry. Therefore, the effectiveness of the proposed algorithm is verified by the workpiece image. The results show that workpiece can be identified more accurately and efficiently through the proposed method.