A Binocular Distance Measurement Method with ORB Features in Region of Interest

Traditional binocular distance measurement techniques require high-precise cameras and have large complexities. In order to improve the accuracy with images taken by normal cameras, we propose a binocular distance measurement method to effective oriented fast and rotated brief features in region of interest (BRI _ ORB). In the method, the binocular camera takes Zhang Zhengyou calibration method to calibrate. Then, the feature points of the region of Interest image are extracted by the ORB features extraction method, and takes RANSAC (Random Sample Consensus) eliminate mismatches. Finally, a nonlinear camera model is established to optimize the measurement distance value. Experimental results show that our method outperforms than the traditional techniques.