摘要
针对无人机机载视频图像由于受到天气或电磁波干扰产生的异常帧引起的压缩码率波动问题,提出了一种基于对图像进行预分析,剔除异常帧后进行压缩编码的方法.对于几种常见的异常帧进行建模,分析了其图像信号特点,结合信号统计特性和其信息熵量度,可靠地检测出视频序列中的异常帧,最后对剔除了异常帧的序列应用H.264标准进行压缩编码.理论分析和试验结果都证明此方法在不丢失无人机机载视频图像序列信息量的情况下,可以保持压缩码率恒定并提高了其压缩效率.
To resolve the video compression bit rate fluctuation due to odd frames generated by the bad weather or electromagnetic （EM） interference in aerial video sequences from unmanned aerial vehicle （UAV） , a new video compression method which analyze video sequences in advance and remove the detected odd frames before video compression was proposed. By constructing models of several popular types of odd frames, the characters of odd frame signals were analyzed. With combination of signal statistics and informa- tion entropy measurements, the odd frames in video sequence were robustly detected. Finally, the video sequence without odd frames was compressed with the new H. 264 standard. Both theoretic analyses and experiments prove that this method greatly improves video compression efficiency of aerial video sequences from UAV and keeps constant compression bit rate, meanwhile does not lose any useful information.
出处
《北京航空航天大学学报》
EI
CAS
CSCD
北大核心
2007年第10期1220-1222,共3页
Journal of Beijing University of Aeronautics and Astronautics
关键词
异常帧
无人机
信息熵
视频压缩
odd frame
UAV
information entropy
video compression
