The automatic sorting system collects the image data by the FVC-310 Vision Controller, and marks the coordinates of the target objects on the line with high speed CCD camera, and calculates the internal parameters and the external parameters of the camera. After obtaining the internal and external parameters of the camera, the position of the target in the robot coordinate system can be calculated from the position of the target in the image, it is connected with 2 robotic arms controllers through 2 network ports respectively.
The traditional machine vision localization technology includes gray matching, contour matching and so on. Gray-scale matching is not accurate and the robustness is poor when there is big variety on the lighting; edge contour matching is more powerful than gray match, it can deal with the situation when the object is partially sheltered or blurred, but it has high algorithm complexity, and it needs the powerful image processing capability of FVC-310.
In this system, a feature matching algorithm based on geometrical shape is used to locate the position and posture of the object. Taking the point set and the directional vector of each point as the measure of feature and similarity, it has better robustness than traditional gray and edge matching, less computational overhead, and better search results when the object in the image rotates, zooms, being sheltered or blurred.