摘要
针对在云天背景下运动红外小目标的分割,对目标在图像序列表现出来的特性进行了分析.根据目标不仅在单帧图像中有自己的灰度分布特性,而且在相邻图像中也具有灰度反相性,提出一种新的分割方法.该方法将单帧处理结果与相邻图像滤波结果相结合,提高了目标的检测概率.实验证明了该方法的正确性和有效性.
For the segmentation of small moving targets in the infrared image sequences, the features of the targets in the sequences were analyzed. Every target has its own grey intensity in a single frame, hut it has reverse phase feature in adjacent image frames. So a new method was proposed to improve the precision of target segmentation. The method combined the results of processing a single image detection probability of targets. Simulation results showed the and filtering the adjacent images, thus increasing the effectiveness and efficiency of this method.
出处
《应用科技》
CAS
2008年第10期5-7,共3页
Applied Science and Technology
关键词
背景抑制
点目标
近邻反相特征
图像融合
background suppressing
point target
reverse phase feature of neighborhood
image fusion
