Vehicle-type identification through automated virtual loop assignment and block-based direction biased motion estimation

Abstract

This paper presents the concept of automated virtual loop assignment and loop-based motion estimation in vehicle-type identification. A major departure of our method from previous approaches is that the loops are automatically assigned to each lane; the size of virtual loops is much smaller for estimation accuracy; and the number of virtual loops per lane is large. Comparing this with traditional ILD, there are a number of advantages. First, the size and number of virtual loops may be varied to fine-tune detection accuracy and fully utilize computing resources. Second, there is no failure rate associated with the virtual loops and installation and maintenance cost can be kept to a minimum. Third, virtual loops may be re-allocated anywhere on the frame, giving flexibility in detecting different parameters.