Easy conductive calibration method for binocular vision system based on collinear image transformation

Abstract

This paper presents an effective technique to calibrate the binocular system. A collinear geometry transformation is derived and introduced into the calibration process. First, we construct a virtual binocular system with every two point pair in the same image, and then an analytic method based on the virtual binocular system is proposed to calculate the internal and external parameters of left and right cameras. The structure parameters are solved from the calibrated extrinsic parameters of each camera. All the parameters of the binocular system are calculated with only one image pair. Furthermore, an iterative refinement by minimizing the metric distance error between the reconstructed point and the real point in a three-dimensional measurement coordinate system is applied to enhance the calibration accuracy. Simulations and real experiments have been carried out. The calibration accuracy is compared with the traditional calibration method. The results show that the proposed calibration methods are efficient in improving the calibration accuracy and simply equipped.