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 ,而且它们的灰度在相邻图像中也具有连续性 ,提出了一种新的分割方法 .该方法把单帧图像的处理结果与相邻多帧图像滤波结果相结合 ,大大提高了目标的分割的精度 .实验证明 。
For the segmentation of the small moving targets in the infrared image sequences, the features of the targets in the sequences are analyzed. It can be observed that the targets have their own intensity distribution in a single frame and their intensities are continuous among several adjacent successive frames. So a new method that integrates single frame processing with temporal filtering is proposed. The method can improve the precision of target segmentation, and is verified by a larger number of results of experiments and comparisons.
出处
《电子学报》
EI
CAS
CSCD
北大核心
2004年第7期1116-1119,共4页
Acta Electronica Sinica
基金
图像信息处理与智能控制教育部重点实验室 (No .TKLJ0 2 0 7)
关键词
背景抑制
目标增强
时间域滤波
图像融合
目标分割
Calculations
Feature extraction
Image segmentation
Signal filtering and prediction
