Abstract: A parallel algorithm for detecting moving targets by imaging sensors and estimating their trajectories is proposed. The algorithm uses the temporal continuity of the smooth trajectories of moving targets and successfully detects and simultaneously tracks all the target trajectories by mapping them from the image sequence onto a single target frame. The algorithm makes use of the spatial consistency of intensity of a target within a short time period, resulting from the continuous 3-D trajectory, to detect the existence of the target. The basic steps of the algorithm, applied to a target frame, involve: 2-dimensional filtering to suppress noise and remove background clutter; target trajectory detection to identify moving targets in the form of linear lines for targets moving with constant speed or parabolas for the accelerating targets using Hough transform; and target following to determine the kinematics of the detected targets and associate targets between successive target frames. The global parallelism of the algorithm stems from the fact that all processing steps involve traditional image processing and pattern recognition techniques for which many parallel implementations exist.!
展开▼
