In order to solve the problem of accurate and real-time positioning and grabbing of workpieces on the automated production line, an improved SURF_FREAK algorithm is proposed for application in workpiece identifying and matching. Firstly, the feature points were extracted by the algorithm using the Speeded Up Robust Features (SURF) algorithm. Then, the FREAK algorithm was added to the mid-distance point pair to describe the feature points. Before the Hamming distance similarity measure, the polar line constraint was added to match the workpiece. The results show that the improved SURF_FREAK algorithm has a great improvement in the recognition speed and matching accuracy of the workpiece compared with the traditional Scale-Invariant Feature Transform (SIFT) and SURF and SURF_FREAK algorithms. The algorithm can be applied to the industrial site to identify the workpiece quickly and accurately, and combined with the binocular technology to complete the positioning of the workpiece. The movement amount of each joint of the robot arm is determined by the kinematics inverse solution, and transmitted it to the controller to capture the workpiece.