摘要
在深入分析空间观测图像中,像素灰度在空间和帧间分布特征的基础上,提出一种基于目标提取的空间观测图像预处理算法,算法包括:噪声分布参数估计、盲元定位与补偿、图像降噪、图像分割以及目标提取等步骤。实验结果表明,该算法能够抑制不均匀星空背景,提取观测图像中的恒星星点和卫星星点,同时使图像信噪比在预处理完成之后有所提高。
The spatial and inter-frame distribution characteristics of pixels in the Space Observation Images （SOl） were thoroughly analyzed. Then, an image pre-processing algorithm is put forward. The algorithm includes： noise distribution parameters estimation, blind pixel calibration, image de-noising, image segmentation and target extraction. The experimental results show that targets in observation image are extracted perfectly and SNR is increased.
出处
《红外技术》
CSCD
北大核心
2008年第9期529-532,共4页
Infrared Technology
关键词
噪声分布参数估计
盲元定位与补偿
图像降噪
图像分割
目标提取
noise distribution parameters estimation
blind pixel calibration
image de-noising
image segmentation
target extraction
