A novel way of understanding for calibrating stereo vision sensor constructed by a single camera and mirrors

As a kind of binocular stereo vision system, stereo vision sensor constructed by a single camera and mirrors (SSCMs) has been paid increasing attention for three-dimensional measurement applications due to its low cost, high efficiency and synchronization. Currently, the existing model for SSCM considers the real camera as two half-Field-of-View (FOV) virtual cameras formed by reflection effects of the mirrors. The calibration methods based on this conventional model are either complicated to operate or deriving non-identical system parameters. To solve the problem, a novel model and the corresponding calibration approach are presented, which take the sensor as an integrated structure with the viewpoint that the camera captures two virtual points formed by the same object point due to the reflection effects of the mirrors. Also, the calibration and the measurement accuracy evaluation functions are discussed. Both simulated and real experiments compare the calibration and the measurement error of the proposed method with the two traditional methods. The results show that the proposed approach has higher measurement accuracy and more robust than the traditional methods, which meet high-precision requirements for low-cost three-dimensional measurement applications.