Nonlinear optimization of optical camera parameter via multi-strategy fusion particle swarm optimization algorithm based on asymmetric circle grid

Authors:

Abstract
Addressing the limitations of checkerboard and circular targets, resulting in low calibration accuracy and eccentricity errors, this study introduces a camera calibration method based on asymmetric circle grid and multi-strategy fusion particle swarm optimization algorithm. To address eccentricity errors by enhancing calibration targets through the refinement and design of asymmetric circle grid, focusing on hardware-based solutions. During target feature point extraction, the imaged circle centers derived from cross-ratio invariance remain unaffected by perspective projection, ensuring accurate feature point localization. After the initial camera calibration, the multi-strategy fusion particle swarm optimization algorithm (MSPSO) is used to enhance the calibration results, improving the method's performance. Enhancements have been implemented in the original particle swarm optimization algorithm (PSO). The non-linear inertia weight and sine-cosine learning factors effectively balance the algorithm's exploration and exploitation phases. Secondly, a fitness-distance balance is utilized to create an alternative learning exemplar instead of the global best position. The population reorganization strategy enhances population diversity and effectively prevents the algorithm from stagnating prematurely. MSPSO is evaluated on the CEC2017 test suite, including function value and convergence curve analyses. MSPSO optimizes camera parameters by minimizing the mean reprojection errors as the objective function. The proposed camera calibration method achieves a reprojection error of 0.086 pixels and a dimensional measurement error of 0.2038 mm, which verifies its practicality and effectiveness.
Funding Information
  • Science and Technology Project of Hebei Education Department (BJ2026230)
  • Doctoral Innovation Project of Tangshan University (2025053ZB)
  • Foundation Project of Tangshan Science and Technology (25130229B)
  • Research Project of Hebei Provincial Department of Education (ZC2026287)
  • National Natural Science Foundation of China (52505567)

This publication has 45 references indexed in Scilit: