Highway Vehicle Ranging System

Ranging system on highway is an important part of ITS (Intelligent Transportation System). It can detect the interval of the car in front. If the interval is less than safe distance, the system will alert the driver to slow down the car.The paper began with analysing some ranging methods wildly used in present. Compared their strongpoints and shortages , bio-stereo vision was chosed to be the basic theory . With Marr's theory on processing of visual information , the ranging system was set up. The whole process of ranging can be devided into three parts. They were setting up the imaging model, character pick-up and depth rebuilding with character matched. They had been presented in details in the paper.In the part of setting up the imaging model, the imaging element had been discussed about their characteristics. The imaging model has been proved to be successful in the experiment. Character pick-up is the key of the ranging system. After summarizing some methods of character pick-up, a more suitable means for ranging system on highway was brought forward. It combined the technology of image smooth, region segmentation and threshold segmentation and picked up the object auto. It has high speed and high veracity in program. With the auto object in the twopicture taken by the two cameras been picked up, the matching technology based on template was put to calculate how much the parallax of the auto in the two pictures . Then camera calibration was done and how far the auto in the front can be got.The arithmetic metioned in the paper had been run in PC with C language . It proved to be a useful one for detect the interval on highway in realtime . At the end of the paper, some assumption was put forward to perfect the system.