基于SIFT和INIR的地表未爆子弹药检测方法研究
Research on surface unexploded submunition detection method based on SIFT and INIR
闫小伟 1石胜斌 1连细南1
作者信息
- 1. 陆军炮兵防空兵学院 高过载弹药制导控制与信息感知实验室,合肥 230031
- 折叠
摘要
采用基于SIFT(尺度不变特征变换)算法的可见光同源图像配准检测方法解决地表未爆子弹药检测问题,并通过试验验证了该方法的可行性,但也发现存在虚警误判的问题.在同源图像配准检测的基础上,提出了基于SIFT和INIR(图像邻域信息重建)的可见光-红外异源图像差异性检测方法,该方法有效解决了误判漏判的问题.对比试验表明,异源图像差异性检测方法的检测精度相较于同源图像配准检测有了较大提升,但也导致无人机载荷平台单次工作时间大幅减少,为下一步地表未爆子弹药目标的准确识别定位提供了有力的技术支撑.
Abstract
The visible light homologous image registration detection method based on SIFT(scale invariant feature transform)algorithm is used to solve the problem of surface unexploded submunition detection.The feasibility of the method is verified by experiments,but the problem of false alarm and misjudgment is also found.On the basis of homologous image registration detection,a visible-infrared heterogeneous image difference detection method based on SIFT and INIR(image neighborhood information reconstruction)is proposed,which effectively solves the problem of misjudgment.The comparison test shows that the detection accuracy of the heterogeneous image difference detection method is greatly improved compared with the homologous image registration detection,but it also leads to a significant reduction in the single working time of the UAV-borne platform,which provides a strong technical support for the accurate identification and location of the unexploded submunition targets on the ground in the next step.
关键词
SIFT算法/图像邻域信息重建/异源图像/未爆子弹药Key words
SIFT algorithm/image neighborhood information reconstruction/heterogeneous images/unexploded submunitions引用本文复制引用
出版年
2024
NETL
NSTL
万方数据