Vehicle-Type Identification Through Automated Virtual Loop Assignment and Block-Based Direction-Biased Motion Estimation

Abstract

This paper presents a method of automated virtual loop assignment and direction-based motion estimation. The unique features of our approach are that first, a number of loops are automatically assigned to each lane. The merit of doing this is that it accommodates pan-tilt-zoom (PTZ) actions without needing further human interaction. Second, the size of the virtual loops is much smaller for estimation accuracy. This enables the use of standard block-based motion estimation techniques that are well developed for video coding. Third, the number of virtual loops per lane is large. The motion content of each block may be weighted and the collective result offers a more reliable and robust approach in motion estimation. Comparing this with traditional inductive loop detectors (ILDs), there are a number of advantages. First, the size and number of virtual loops may be varied to fine-tune detection accuracy. Second, it may also be varied for an effective utilization of the computing resources. Third, there is no failure rate associated with the virtual loops or physical installation. As the loops are defined on the image sequence, changing the detection configuration or redeploying the loops to other locations on the same image sequence requires only a change of the assignment parameters. Fourth, virtual loops may be reallocated anywhere on the frame, giving flexibility in detecting different parameters. Our simulation results indicate that the proposed method is effective in type classification.