A Study on Traffic Parameter Extraction at Various Intersection via Image Processing

• Main Authors: Yu-Shiang Lin, 林于翔
• Other Authors: 張堂賢
• Format: Others
• Language: zh-TW
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/22604075464513084630

碩士 === 國立臺灣大學 === 土木工程學研究所 === 97 === The traffic parameter at intersection is very important issue to traffic control. It is not only a basis of the road design, but a setting standard of the traffic signal. In recent years, more and more people research about how to develop an image detector system to collect traffic parameters. Since there is much information in the image, we can extract traffic parameter by image processing and computer vision technology. The study of image detector on highway had been developed mature enough that it could extract traffic parameter with highly accurate rate. However, the researches about data collection at intersection are rare. The possible reason may be that intersections are more complex than highways; vehicles are difficult to be tracked well at intersections. In Taiwan, there are many motorcycles and pedestrians at intersection. Because of some characteristics of it, motorcycles are difficult detect at intersections. It becomes a challenge for image detector system to have reliable ability of vehicle detection. This study is mainly composed of five stages: (1)pre-processing, (2)foreground segmentation,(3)shadow removal, (4)vehicle tracking and (5)traffic parameters extraction. The pre-processing is developed to obtain the information of road geometry and calibrate the camera. After the preprocessing is done, the foreground segmentation and shadow removal continue to segment the moving vehicles from the input images. According to the results, the average success rates of different vehicles counting are higher than 80%. Moreover, it shows that this system is capable of successfully extracting the traffic parameters, including trajectory of the moving vehicle at road and various intersections.