Abstract: According to the disadvantages of conventional binocular vision sensor such as big size, low measuring efficiency, high cost, no synchronal image collection and so on, a structural model of single camera binocular vision sensor using mirrors is established. An optimization structure design of single camera binocular vision sensor using mirrors is proposed. A single camera and the mirror optical system which consist of two angled mirrors constitute the sensor. Two virtual images reflected by two mirrors of one object are acquired in a single shot by the camera set in front of the mirror optical system. The conjugate points from the left part and the right part of the image is corresponding to the same object. It is equivalent to the traditional binocular vision sensor in which the same object is obtained by two cameras. Field of view is related with the size of mirrors, the angle between two mirrors and the distance between camera and mirrors. Thus the structure of the sensor can be determined according to the requirement of field of view. The effect of the geometry on the measuring accuracy is analyzed according to the relationship between the structure parameters and the position of measurand. The reasonable range of parameters is estimated by simulation. Based on the proposed principle and procedure, structure design can be instructed by the theory under different circumstances. The research shows great significance in application.

Key words: Accuracy analysis, Binocular vision, Mono-camera, Structure design