APPLIED LASER, Volume. 41, Issue 4, 860(2021)
Optimized PSO Space High-precision Line Structured Light Direct Calibration Method
The three-dimensional calibration of the line structured light measurement system is a necessary part of the three-dimensional measurement. Its calibration accuracy directly determines the accuracy of the three-dimensional measurement. The calculation process for the current line structure cursor positioning is complicated, limited, high cost, low precision, etc. Problem, this paper proposes a direct calibration method of spatial high-precision line structured light. Design a special-shaped calibration block, remove the noise points and boundary points on the line laser stripe of the calibration block through the DBSCAN algorithm, then use the improved PSO algorithm to optimize the objective function of the shortest path optimization, extract the characteristic calibration points on the line laser stripe, and establish the pixel coordinates The mapping relationship table with the world coordinates, so as to directly obtain the world coordinates corresponding to the pixel coordinates of the point to be calibrated, or calculate the relationship conversion matrix through the three characteristic calibration points closest to the point to be calibrated to obtain the world coordinates of the point to be calibrated. Experiments show that the optimized PSO algorithm′s optimization ability has increased by 27.23%, the convergence speed has increased by 66.67%, and the extraction error of the feature calibration point is about 0.4 pixels. After the linear structured light measurement system is calibrated, the direction along the x direction The average absolute measurement error is 0.0156 mm, and the average absolute measurement error along the z direction is 0.0174 mm.
Get Citation
Copy Citation Text
Wang Hua, Guo Yulong, Zhang Shuang, Gao Jingang, Liu Xiaofeng. Optimized PSO Space High-precision Line Structured Light Direct Calibration Method[J]. APPLIED LASER, 2021, 41(4): 860
Received: Oct. 12, 2020
Accepted: --
Published Online: Jan. 10, 2022
The Author Email: Hua Wang (1336528482@qq.com)