Calibration of Binocular Vision Sensor with Large Field-of-View Based on Two Small Planar Targets

Abstract

The current calibration methods for binocular vision sensor (BVS) with large field-of-view (FOV) require large scale targets, which are hard to be manufactured, inconvenient to be posed and with low accuracy. In view of the above problems, this paper presents a method for calibrating BVS with large FOV based on two small planar targets connected rigidly. Firstly, the two small planar targets are approached nearly to the BVS, and each camera shoots the small target corresponding to it respectively. Then, the location uncertainty of the image feature points of targets is determined. The location deviation of image feature points and the cameras intrinsic parameters are derived by non-linear optimization. Finally, according to the constraint that the relative position of the two planar targets is invariable, the optimal solution of the transformation matrix can be obtained by non-linear optimization. Experiment verifies the effectiveness of the proposed method.