无人机序列图像压缩方法研究

Methods for compressing sequence image of unmanned aerial vehicle

根据无人机的运动特点,求出序列图像之间的重叠区域,将运动序列图像转变成静态图像,然后以类EBCOT算法完成拼接图像的压缩.压缩的实验结果显示,该方法在压缩效率和运行时间上均优于H.264等视频压缩标准,可较好地满足无人机图像的传输需求.

On the basis of the analysis of the video compression standard, a new and more efficient compression method was put forward in accordance with the characteristics of sequence image of unmanned aerial vehicle （UAV）. In consideration of the movement characteristics of UAV, the overlapped areas in the image sequence were deduced and a moving sequence image transformed to a still image. The compression of the mosaic image was completed by the EBCOT-like algorithm. The compression experiments results showed that this method is superior to the video compression stands such as H. 264 standards, etc. in terms of compression efficiency and running time, meeting the transmission requirements of the image of UAV.