海天复杂背景下红外目标的检测跟踪算法  

IR Target Detection & Tracking Algorithm Based on Sea-sky Background

作　　者：苏秀琴[1] 梁金峰[1,2] 陆陶[1,2] 杨露[1,2] 

机构地区：[1]中国科学院西安光学精密机械研究所,西安710119 [2]中国科学院研究生院,北京100049

出　　处：《光子学报》2009年第5期1309-1312,共4页Acta Photonica Sinica

摘　　要：在分析海天复杂背景下红外目标图像特征的基础上,提出适合该环境的红外目标检测算法.该算法采用行均值相减的方法抑制海平面非线性温度场的影响,并进行中值滤波处理.对于更加复杂的环境,选用数学形态滤波法抑制背景中的大面积云团或海浪,从而确定出目标区域来进行目标图像的分割及增强.同时,综合使用图像捕获区域指定、运动目标检测法、弱目标的增强提取、记忆外推功能、数据融合加权跟踪方法,来保证在海天复杂背景下红外目标的可靠跟踪.实验表明,该算法能较好地处理海天复杂背景下红外目标的检测,且算法易于硬件实现,提高目标检测的实时效率.Based on the characteristic analysis of infrared image target under the sea-sky background, a target detecting algorithm was proposed based on the IR image. With line average subtracting, the nonlinear influence in temperature field could be inhibited. Then with mathematical morphology, it could inhibit the cloud and ocean wave in the background. After ascertaining the region of the target, the targets could be separated off with image segmentation and enhancement. For tracking reliably,moving target detection , small ＆ weak target detection,ballistic trajectory forecast tracking , tracking with weighing etc algorithm were introduced. The experiment results show that the algorithm can efficiently inhibit the influence and detect the target under the sea-sky background.

关 键 词：红外图像 目标检测 行均值相减 数学形态学 跟踪 

分 类 号：TP751.1[自动化与计算机技术—检测技术与自动化装置]