With the upgrading of China's industry, the assembly line operation of the factory has higher requirements for automation and intelligence. For industrial robots, machine vision to guide operations is becoming more and more popular, and the demand for three-dimensional vision guided robots is increasing. Three-dimensional vision can solve problems that cannot be solved by traditional two-dimensional vision, and has the potential to replace traditional two-dimensional vision. At present, there are problems such as low positioning accuracy, poor robustness, and difficulty in adapting to the positioning requirements of complex workpieces with curved surfaces due to the lack of 3D information, and 3D vision can well cope with the positioning problems of such workpieces. In this study, the traditional ICP registration algorithm is improved to meet this demand. Finally, compared with the traditional ICP registration algorithm, the registration effect and speed of the improved ICP registration algorithm are better than before