Abstract
In order to prevent traffic accidents on expressways, it is necessary to give drivers information regarding existing danger in advance. Many different sensors have been developed for detecting danger on expressways. In current traffic monitoring systems, visual sensors are often applied, and humans directly survey the expressways through monitors. However, the performance of current systems usually fails when undesirable images are generated under a poor environment such as brightness changes and low contrast. Avehicle detection method was developed based on Rotation Vector Matching,and it was confirmed that this system works effectively even when operating in a poor environment. For the covering abstract see ITRD E134653.