CAJ | 학술논문

红外舰船目标检测是舰船目标识别与跟踪的重要基础。针对典型复杂海天场景下的红外舰船目标检测，本文提出了一种基于多尺度局部边缘梯度统计特性的目标检测算法。算法改进了传统的基于ROI的红外舰船目标检测方法，利用多尺度局部边缘梯度的统计特性，自适应的检测出完整的舰船目标。实验结果表明，该算法能够有效准确地检测出多种复杂海天场景下的红外舰船目标，且算法简单易于实现。
홍외함선목표검측시함선목표식별여근종적중요기출。침대전형복잡해천장경하적홍외함선목표검측，본문제출료일충기우다척도국부변연제도통계특성적목표검측산법。산법개진료전통적기우ROI적홍외함선목표검측방법，이용다척도국부변연제도적통계특성，자괄응적검측출완정적함선목표。실험결과표명，해산법능구유효준학지검측출다충복잡해천장경하적홍외함선목표，차산법간단역우실현。
Infrared warship target detection is an important foundation of the warship target recognition and tracking. For infrared ship target in typical complex sea-sky environment, this paper proposes a target detection method based on multi-scale edge gradient. Algorithm improved the traditional infrared ship target detection method which based on ROI(Region of Interest) and using statistical properties of multi-scale edge gradient adaptive detected complete ship targets. Experimental results indicate that the algorithm not only can effectively and exactly detect the infrared warship targets in different complex sea-sky background, but also shows simple and easy to implement.