Automatic TV Image Tracking Using Advanced Positive Frame Difference Predicted Window Technique.

TV image tracking is a subject of increasing interest due to the progress of the new technologies and fabrication of high resolution long range cameras and fast microprocessors. A new technique for TV image tracking; advanced predicted window technique (APWT), is proposed. In this technique the tracking window is positioned at an advanced (predicted) position of the target moving center in a noisy background. In addition, the advanced positive frame difference predicted window technique (APFDWT) is presented, and its mathematical formulation ( including separation of moving target from noisy background, extraction of dominant features of moving target, and selection of target to be tracked ) is derived. The proposed system has the ability to identify and track close distinct targets flying in noisy stationary ambient-ground. The evaluation of target center, orientation angle, and velocity parameters are introduced to improve the tracking process.
The proposed system overcomes most of the disadvantages of the traditional TV tracking systems such as : initial manual location of the window around the selected target, initial adjustment of the window size, tracking failure due to the movement of other targets across the window, and the increased processig time due to the use of micro image processors in case of multi-target tracking. Simulation results show that the proposed tracking system is more stable and reliable .