Zernike moment based time sensitive targets detection and recognition

摘要

The ground time sensitive targets have high mobility. It is hard to use traditional template matching method to detect and recognize the targets since their attitude angles are not known. To make matters worse, the complicated background and grayscale reversal always cover the target like airplane and warship. This paper proposes a method based on 2D Zernike moment and Adaboost algorithm to detect and recognize the time sensitive targets. The experimental result indicates that the method is efficient.