A binocular vision system is composed of two sets of cameras, light sources and image processing industrial personal computers. It uses binocular vision technology to implement workpiece positioning in 6 dimensions (position and angle in X, Y and Z directions) and guides the accurate robotic grabbing upon correction. It can be widely used for material handling, assembly, loading and unloading in various production lines. The camera may be fixed or installed on the robot gripper. The global field of view and strong scalability of the cameras enable easy addition of vehicle models to the production line.
Principles of vision:
The two cameras for binocular stereo vision collect workpiece images from different poses, calculate the depth information based on the image parallax corresponding to the same measured feature point on the workpiece, and conclude the spatial pose of the workpiece from the three-dimensional position information of multiple measured points.
