论文部分内容阅读
针对高空无人机航拍序列图像的实时图像压缩问题,设计了一种基于单应变换的可精确定比的高压缩性能低计算复杂度航拍序列图像压缩算法.该算法使用机载惯导数据进行编码帧组划分并获得中间帧相对于首尾帧的初始单应变换,进而通过多子区匹配实现对单应矩阵的快速精确求解,然后通过单应变换来实现对中间帧的整体运动估计和补偿,最后用JPEG2000算法对帧组数据进行整体压缩.实验结果表明,相对于传统算法,该算法具有计算复杂度低、压缩比高、保真度好、可精确定比压缩的优点.
In order to solve the problem of real-time image compression of high-altitude UAV aerial sequence images, an image compression algorithm with high compression performance and low computational complexity of aerial sequence based on homography transformation is designed. The algorithm uses on-board inertial navigation data Coding frame group division and obtain the initial homography transformation of the intermediate frame relative to the first frame and the last frame, and then realize the fast and accurate solution of the homography matrix through the matching of the multiple sub-regions, and then realize the integral motion estimation and compensation of the intermediate frame by the homography transformation Finally, JPEG2000 algorithm is used to compress the entire frame data.The experimental results show that the proposed algorithm has the advantages of low computational complexity, high compression ratio, good fidelity and accurate compression compared with the traditional algorithm.