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 spatial frequency spectra. Frequency distribution depends on the road surface type. Ground velocity is shown by a filtered moving image. Experimental results show that five road surface types were recognized in cold districts. Both road surface condition and vehicle velocity can be detected by a noncontact sensor; therefore, the sensor is useful for active safety systems.