Radial Distortion Correction of Grid Images Based on Support Vector Machines for Regression Using Particle Swarm Optimization Algorithm

Distortion correction is a key technology in the filed of machine vision. A radial distortion correction method based on support vector machines for regression is proposed and studied thoroughly. Firstly, a grid image with radial distortion is segmented. Secondly, the segmentation image is refined to get a skeleton image, and intersection points of the grid are extracted. Thirdly, the actual distortion locations of these intersection points and their corresponding ideal distortionless locations are combined to construct a training set and test set of support vector machine, and a particle swarm optimization algorithm is used to determine the optimal parameters of support vector machine. Finally, a support vector machine for regression is utilized to correct the radial distortion of the grid image. Some intuitive and persuasive simulation experiments are carried out by using the proposed distortion correction method. Experimental results demonstrate that the proposed method can correct the radial distortion of the grid image. The proposed distortion correction method can be used in the fields of image measurement, three-dimensional reconstruction, collaborative augmented reality and teleoperation robots.