The spatial filtering method is useful for the non-contact velocity measurement of moving object on which a random surface pattern or texture exists. A remarkable feature of this method is the simple structure and rapid response of optical parallel processing. However, there remains a serious problem that no signal is obtained if the characteristic frequency is not contained in spatial frequency components on the object's surface. In order to overcome this problem, pitch of the spatial filter must be adjusted to match the spatial frequency characteristics of the object's surface. But it is difficult to change the shape of the parallel slit reticle of the conventional spatial filter.
In this paper, we propose an electronic spatial filter utilizing an imaging device and full digital circuits. Various shapes of weighting function of spatial filter can be easily realized by this approach. A dual spatial filter having orthogonal sinusoidal weighting function is implemented. Its output can be utilized to sample a spatial frequency component out of the object surface.
An intelligent velocity sensing system is fabricated. It can evaluate the matching of filter performance with its object and adjust the reticle pitch automatically to meet change in the characteristics of the object surface. Features of this system is proven by experiments.