Study On Container Position System Based On Computer Vision

The containerized transportation is one of the three symbols of modern trade development. Container loading and unloading represent the most advanced productivities of shipping in the world. The efficiency of handling container ships in harbors directly impacts the competition abilities of shipping companies and harbors in the world market. How to realize the automation of container loading and unloading has been becoming one of the hottest research areas in logistics. Currently, using computer vision technology is the direction of container recognition and position in automatic container loading and unloading operations. The cost and the level of risk should be high to research the feasibilities of applying computer vision in container loading and unloading in an industrial scene, because a container loading and unloading system is very complex and the computer vision technology involves multiple disciplines.This thesis proposes a method of constructing a simulation system of container loading and unloading to invest the feasibilities of applying computer vision technology in automatic container loading and unloading operations. The research works have been conducted in the following areas. First, the simulation system is designed and constructed according to the real container loading and unloading environment. Then, the communication protocol between a PC and the main movement controller of the system is composed. The system's software is also developed. Finally, the key research works of the system are conducted, including the algorithms of image processing, the building of a camera model, and the calibrations of the camera and the system's mechanical parameters. Thus, the recognition and position of simulated container are realized.In this thesis, theoretical analyses and the experimental studies are conducted on the simulation system of container loading and unloading, including working principles and constructions of system software and hardware, image processing, and working principles of the computer vision system. The simulation system is constructed and its design goals are reached. Furthermore, some new methods are proposed for step motor driving and the pre-process of image data. The establishment of the simulation system of container loading and unloading provides a foundation to the further studies on computer vision technology in automatic container loading and unloading. The simulation system can work as a platform for research on increasing the efficiency of container loading and unloading in harbors.