论文部分内容阅读
针对复杂背景下红外目标检测问题,提出一种联合熵流与边缘算子的红外运动目标检测算法.该算法分4个步骤检测目标:首先将熵流作为一种图像运动描绘子,通过熵流确定含目标运动的区域;其次运用Canny算子检测含背景和目标的边缘;然后联合两者捕获背景成分急剧降低、近似精确的目标边缘图像;最后计算边缘图像聚类中心,依据亮度、熵流约束进行区域生长,识别红外目标.对于海空背景下的红外目标图像,实验结果表明该方法能准确地检测、定位目标.
In order to solve the problem of infrared target detection under complex background, an infrared moving target detection algorithm based on entropy flow and edge operator is proposed. The algorithm detects the target in four steps: firstly, entropy flow is treated as an image motion descriptor, Then the region containing the target motion is determined. Secondly, the edges of the background and the target are detected by using Canny operator. Then the combined background and the target are sharply reduced to approximate the exact target edge image. Finally, the edge image clustering center is calculated, Constraints of regional growth and identification of infrared targets.Experimental results show that the method can accurately detect and locate the target of infrared target images in the sea and sky.