Research on calibration feature optimization and adaptive visual parameter adjustment for complex grating measurement

Abstract

This paper presents a novel method for the intelligent adjustment of vision parameters in structured light camera calibration under complex light conditions, aiming to enhance accuracy and reduce interference from human and external factors. Firstly, a self-learning weight calibration feature extraction model (SLWFE model) is developed to solve the coupled interference problem of calibration feature extraction. Secondly, we analyze the influence of focal length on structured light phase-height mapping accuracy and construct a grating calibration characteristic gradient filter function. The focus confidence evaluation model of calibration image is proposed, to realize the accurate calculation of optimal exposure time and lens ideal focus position, leading to the development of the grating calibration image characteristics optimization algorithm (GCICO). Finally, an intelligent parameterization device and control system were created, integrating the algorithm for experimental verification, achieving an average reprojection error of 0.018 pixels and an improvement of 70.49% over traditional methods.