Abstract:The most critical step of the line structured light three-dimensional vision measurement technology is to extract the center line of the laser stripe in the structured light image. In order to solve the problems of complex background information, uneven distribution of laser intensity, wide difference of width of different parts of light band and laser fringe fracture in dynamic environment, a method of extracting laser fringe center line suitable for dynamic environment was proposed in this paper. First, the region of interest (ROI) is extracted by image preprocessing and adaptive clipping algorithm. Secondly, the laser fringe region was segmtized by the improved gamma correction (IGC) and the improved variable threshold Otsu (IVT-Ostu). Then, the initial center line of the laser fringe is extracted by the two-dimensional gray barycentric method (TD-GBM). Finally, a quadratic optimization algorithm is proposed to optimize the initial center line and extract the center line of laser fringe accurately. Experimental results show that, compared with gray gravity center method, Steger method and other algorithms, the proposed method is less affected by background interference and laser fringe quality, and can extract the center line of laser fringe more accurately in a variety of complex situations, meeting the requirements of high accuracy, strong stability and good real-time performance.