Three-dimensional surface reconstruction based on binocular vision

Three-dimensional reconstruction is an important application aspect of machine vision. The relevant method is proposed to complete the three-dimensional reconstruction of the target object in this paper. Firstly, the non-linear optimization algorithm is proposed to calibrate the vision system whose the parameters are used to correct the target object. Secondly, the target object is extracted based on the combination of gray-scale segmentation and feature segmentation. Then, gray-scale histogram matching is added to improve accuracy of the normalized cross correlation (NCC) matching algorithm. Finally, the three-dimensional coordinate of target object is calculated. The suitable camera and lens are selected and a binocular vision platform is built to verify the feasibility of the proposed algorithm by experiments.