Camera Calibration Method Based on Sine Cosine Algorithm

Abstract

Aiming at the problems of traditional camera calibration method, such as sensitivity to the initial values of camera model parameters and unstable calibration results. This paper proposes a camera calibration method based on sine cosine algorithm. After obtaining a certain initial value by Zhang's camera calibration method, use the sine cosine algorithm (SCA) to form the initial population in the field near the initial value, and perform iterative optimization. The average error between the actual projection point and the calculated projection point is the accuracy criterion. Using the volatility and periodicity of the sine function and cosine function to search and iterate, so that the solution can be oscillating towards the global optimum and achieve the purpose of optimization. Experiments have proved that the adaptive parameters and randomness parameters in the algorithm better balance the exploration and development capabilities of the algorithm. The improved algorithm has fewer parameters, simple structure, easy implementation, and fast convergence speed. The experiment proves that the camera calibration accuracy is effectively improved.