摘要
为了解决机载红外预警探测系统检测地面运动点目标时的结构化背景抑制,提出了一种基于非下采样Contourlet变换的新算法.算法采用非下采样Contourlet变换对原始图像进行多层分解,然后对低频子带和高频子带采用不同的方法处理,最后对各子带进行重构即可得到背景抑制后图像.与数学形态学Top-hat算法比较,实验结果表明本文所提算法能有效地抑制图像背景,从而较好地提高图像的信噪比和对比度.
A new infrared ground point background suppression algorithm based on the Nonsubsampled Contourlet Transform （NSCT） is presented. Raw image is decomposed using NSCT, different operators are adapted in different decomposition scales and directions, and background suppression image is obtalned by NSCT reconstruction. When compared with conventional mathematical morphology Top-hat algorithm, several groups of experimental results demonstrate that the presented algorithm can suppress background effectively and the signal-noise ratio and contrast of image is improved efficiently.
出处
《光子学报》
EI
CAS
CSCD
北大核心
2009年第12期3318-3321,共4页
Acta Photonica Sinica
基金
教育部科学技术研究重点项目(108114)资助
