| Popis: |
Robust and efficient air infrared small target detection is widely used in reconnaissance and early warning system. However, there exist two critical difficulties, i.e., the limited detection range and the poor detection performance of single infrared sensor. To overcome those difficulties, in this paper, we propose a novel cooperative detection method of multiple infrared sensors for air small target. Specifically, a multiscale block local contrast measurement method is introduced to achieve the target detection of each infrared sensor. Afterwards, a double-layer ring grid corner feature extraction method is proposed to achieve the image stitching accurately between multiple infrared sensors. In addition, the gray correlation analysis with position constraint based feature level fusion method is exploited to match targets in different sensors. Experimental results on testing image sequences demonstrate that the proposed method can improve the detection rate and reduce the false alarm rate effectively on the condition of limited detection range of each single sensor. |
