The human centered ITS view aid system is an advanced driver assistance system aiming at helpful, appropriate and not annoying display of information and warning for any driver. It consists of on-board sensing systems, driver monitoring systems, an inter-vehicle communication system, and an information and warning display system. Detection of vehicle driving environments is a function in the on-board sensing system, and the detection of it is necessary to secure transport facilities from accidents and to keep the performance smooth. The road surface condition is one of the important factors toward detection of vehicle driving environments. So, this paper relates to the study of a road surface condition detection technique. The information related to the water and/or snow on the road, and the fusion and freezing caused by changing the temperature is closely connected with the changes of the road conditions. Therefore, the road condition detection can be accomplished by extracting the features quantitatively about these factors and integrating them appropriately. These features are extracted quantitatively from the images taken by TV camera attached at a rearview mirror. The paper examines a technique to detect the water on the road surface at the beginning. The extraction of the water on the road is made with the technique that gets the property of the polarization from the image. Features related to water were extracted by the ratio of horizontal polarization image intensity to vertical polarization image intensity for each pixel. The paper describes how field tests were conducted in order to verify the detection ability of the road surface water while running the expressway at 100km/h on the average and got favorable results.
Abstract