A self-adaptive calibration method for mine-used belt detection with binocular visio

For traditional complex camera calibration methods with low precision used for mine-used belt longitudinal tear detection, an efficient self-adaptive calibration method was proposed for binocular vision detection. A mathematical model of camera and fundamental principle of binocular vision were analyzed, and non-linear distortion parameters were introduced into the linear model. The coordinate values of feature corner points extracted from fused image combining belt image with a 7×7 matrix model are substituted into the matrix constraint equation. Furthermore, internal and external parameters, structure parameters and distortion parameters decomposed of the equation are non-linearly optimized to be more accurate. Finally, overall calculated parameters are compared with the ones obtained by Faugeras self-calibration method using Bayesian estimation error. The experimental results show that the method has high accuracy and reliability.