Automatic Identification System Based On Video Models To Achieve

The thesis is supported by the Innovation Foundation of Excellent Intellectual in Henan Province.To date Intelligent Transportation System (ITS) has become one of the most popular worldwide research fields in traffic technology. The research topics acknowledged internationally include advanced traffic management system, advanced traffic information service system, commercial vehicle transportation and management system, electronic toll system, public transportation management system, emergency management system, and advanced car monitoring system. As a representation of electronic automatic toll system, the Non-Stopping Electronic Toll System (NSETCS) based upon non-contacting Integrate Circuit (IC) card is a useful method to solve this problem because it can effectively improve the transportation and service of highways.Since car model is the essential base to determine the highway toll, Car Model Identification System (CMIS) is an important part of The NSETCS based on non-contacting IC card, which normally adopts information fusion technology to make a decision. The system fuses the information of the Car Model exported by CMIS, license plate exported by License Plate Recognition System, and the data saved in the non-contacting IC card to obtain toll decree and hence greatly improves the reliability of the system. The thesis presented a real-time CMIS based on the video image acquired from video cameras by employing moving object detection algorithm, feature extraction and pattern recognition algorithm.The main works in this thesis are summarized as follows:1. Designed and implemented a Video Image Processing System based on DirectShow, and presented the kernel software code.2. Studied the existing digital image processing technology and algorithms, and applied them to Car Model Recognition System to suppress all kinds of noise.3. Reviewed the main moving object detection algorithms and their advantages and disadvantages. Developed a new algorithm for the detection of high-speed moving objects based on picture blocks and HVS color space of image differences techniques.4. Feature extraction. Seven car image invariant moments are extracted from the car...