Target location of Loitering Munitions based on image matching

Abstract

This paper presented a method for high accuracy ground target localization by using Loitering Munitions (LM) equipped with a video camera sensor. Firstly, based on tray tracing ground target geo-location was used for comparison against our approach, and the calculation from pixel to world coordinates can be done. Then, in order to remove small details hough transform was used to image alignment and a median filter was applied which are visible from the sensed image but not visible from the reference image. Finally, based on bifurcation extraction a novel edge detection method and an image matching algorithm were proposed. This method did not require accurate knowledge of the aircraft position, attitude and high performance sensors. Therefore, it is especially suitable for LM which did not have capability to carry accurate sensors due to their limited play weight and power resources. The results of simulation experiments and theory analyzing demonstrate that high accuracy ground target localization is reached also using low performance sensors, and achieved the real time need. Additional, comparing with the matching algorithm based on normal cross correlation, the proposed algorithm has lower computational burden and faster processing speed, which can meet the real-time requirements for scene matching.