The improvement and application of intelligence tracking algorithm for moving logistics objects based on machine vision sensor

Logistics object tracking is one of the key technologies of intelligence logistics management system (ILMS). Using machine vision sensor technology, the paper introduced the application method of machine vision sensor technology in intelligent logistics system, and proposed an improved and simple covariance matrix algorithm to detect and track moving logistics objects. Against the features and technical difficulties of moving logistics objects detection, the covariance matrix algorithm was applied to detecting and tracking of logistics objects, and against the shortcomings of covariance matrix algorithm in the process of detecting and tracking of moving logistics objects, the paper presented a method of logistics objects path prediction, and template image dynamic selection and adjustment. Experiments show that the method can effectively apply the improved covariance matrix algorithm to the detecting and tracking of moving logistics objects, the method can not only adapt quickly to pose and scale variations of logistics objects, but also track accurately and continuously those temporarily occluded logistics objects, the improved method has good robustness. The method provides a new solution of the detecting and tracking of moving logistics objects.