摘要
为获得高质量视频输出序列,电子稳像技术常被用来去除成像设备所摄取的图像序列中的随机抖动。本文首先介绍了电子稳像的国内外发展现状。然后,从电子稳像技术中的运动估计模块出发,描述了国内外近年来出现的直接求取运动估计算法以及各种基于特征点匹配的运动估计方法,并对运动校正和运动补偿模块的算法进行了评述。最后综合分析了稳像算法的难点以及未来发展趋势,提出基于特征匹配的电子稳像技术的发展方向。
Electronic Image Stabilization ( EIS ) technology usually is employed to remove the random jitter generated by the imaging equipment and to stabilize its video output .Firstly, the development status of EIS is introduced .Then the motion estimation module is studied in details , which is one of the most important mod-ules of EIS .Both the direction motion estimation algorithms and the feature-points-based matching algorithms are discussed.Furthermore, the motion correction and compensation algorithm are presented .Finally, accord-ing to the difficulties and the future trends of the EIS , the developing direction for feature matching algorithms of EIS is proposed .
出处
《中国光学》
EI
CAS
2013年第6期841-849,共9页
Chinese Optics
基金
国家自然科学基金资助项目(No.11071103)
关键词
电子稳像
运动估计
特征匹配
卡尔曼滤波
electronic image stabilization
motion estimation
feature matching
Kalman filter