Design Of Intelligent Object Tracking Luggage Cart System Based On Binocular Vision

With the continuous development of computer vision technology, binocular vision technology can be applied in many fields. Based on characteristics of binocular stereo vision system, this paper designs a new binocular vision-based intelligent tracking of luggage cart system. The research includes the calibration of the camera, the feature extraction of the detected moving target, stereo matching and 3D reconstruction.The camera imaging model and the principle of binocular vision system are described carefully in this paper. The CMOS cameras are used to build the binocular vision system. After the comparison of the two-step method and Zhang's method, the Zhang's method is chosen to implement the camera calibration. The commonly used adjacent frame difference method is used to detect the moving target, the centroid two-dimensional coordinate of the moving target is extracted and recorded, and the pre-match model is built based on the differential image of the moving target. Then edge features of moving targets are extracted by the Canny edge detection algorithm, and candidate matching points are chosen from edge feature points. Detect and match from the same point of principal and subordinate view field through interaction correspond constraints and epipolar constraint equations. Finally the three-dimensional reconstruction model of the binocular vision system is used to calculate the three-dimensional coordinates of matching feature points to track the object. Based on the characteristics of the binocular vision system, system components of the smart luggage cart and the system workflow are introduced, and the target recognition and tracking algorithms and the control strategy are described carefully. Matlab simulation shows that the tracking effect of moving target and the extraction of depth information are both well.