Real-time Detection for Critical Parts of Aerial Reconnaissance Balloons

HanYu Hong

Abstract:

Detecting and striking the critical part
of balloons that is an important approach to counterattack the reconnaissance. The
existing algorithms of the image target detection task is nor able to obtain
the high accuracy and real-time performance in a time, since the critical part
of balloon is weak and not easy to segmented. In this paper., a real-time
algorithm based curvature feature in polar coordinate system is proposed to
detect the critical parts of reconnaissance balloons. We divide our proposed
method into three steps: Firstly, the image is subjected to grayscale
projection and third-order post-difference calculation. Then the polar
coordinates of the balloon are transformed and the balloon boundary is
extracted. Finally, according to the curvature of the balloon boundary, the coordinates
of the critical parts are computed. The strategy of the proposed method is to
adopt the boundary features of balloon instead of general time-consuming image
operations (e.g. region labeling, matching) to capture the target part. The
experimental results show that the proposed method performs the high accuracy results
and real-time detection. Our proposed method achieves a processing speed of 200
frames per second on DSP (TMS320C6678) while maintaining a state-of-art
detection accuracy (>94%), which overcomes the existing comparison algorithms.