Road surface recognition sensor using an optical spatial filter

Abstract

A road surface recognition sensor has been realized by using an optical spatial filter. The sensor transforms a road surface image into a spatial frequency spectrum. The frequency distribution depends on the road surface type. Ground velocity is shown by a filtered moving image. In order to develop a stable system with this method, we have constructed a system which detects normal reflection for the recognition of road surface and diffuse reflection for the acquisition of vehicle velocity, alternately. Experimental results show that seven different types of road surface were correctly recognized in cold districts. Both road surface condition and vehicle velocity can be detected by a noncontact sensor. This sensor is convenient for the information systems of road surface condition.