摘要
本文首先介绍了2种主要透镜畸变及摄像机标定方法;然后结合摄像机内部参数和视频帧率推导出关键帧航向重叠度计算公式;在此基础上实现了指定航向重叠度关键帧从文件源或实时源中的自动提取,文件源中采用定位关键帧自动提取算法,实时源中采用定时自动提取算法;最后利用得到的标定参数完成关键帧影像矫正。实验结果表明本文采用的算法能够高效、准确地得到矫正好的指定航向重叠度关键帧。
In the paper, first, two main lens distortions and camera calibration were introduced. Then the formula of key frames forward overlap was developed within intrinsic parameters of the camera and video frame rate. Then key frames were automatically extracted from the file or real time source. Automatie extraction of timing from real time source was processed and key frame positioning method was used for extracting key frames from file source. Finally key frames were corrected with the calibration parameters. Real data was used to test the developed method, and results showed that the technique was efficient and exact.
出处
《测绘科学》
CSCD
北大核心
2012年第5期110-112,115,共4页
Science of Surveying and Mapping
基金
国家"863"项目(2008AA121305)
国家"863"项目(2009AA121403)
国家自然科学基金资助项目(40901221)
中国博士后科学基金资助项目(20090450182)
关键词
视频流
关键帧
自动提取
重叠度
video streaming
key frame
automatic extraction
overlap degree