摘要
针对动态图像的遮挡缺损问题,提出了基于图像块的帧间预测算法,可以检测出序列图像中的遮挡缺损区域;利用帧间灰度相关性修正补偿算法进行遮挡缺损区域补偿,利用修正的像元对缺损区域进行补偿能够克服传统插值补偿算法补偿痕迹明显的缺点。仿真结果表明:该算法补偿处理区域与邻域灰度过渡平滑,消除了补偿痕迹,补偿效果比较理想。
To solve the image obscuration in dynamic images, an image block algorithm based inter-frame predictor is presented to detect missing image region in sequence images and a pixel inserted compensation algorithm based on gray correlation is given. The inter-flame predictor based on image block algorithm can be used to detect the missing image region in sequence images, which is caused by a partially obscured CCD. A method using modified pixels with inter-frame gray correlation is presented to compensate the missing image region. The simulation results show the compensation algorithm is effective.
出处
《应用光学》
CAS
CSCD
北大核心
2009年第4期601-604,共4页
Journal of Applied Optics
关键词
遮挡图像
图像补偿
帧间预测
灰度相关性
obscured image
image compensation
inter-frame predictor
gray correlation