基于小波的感兴趣区域无人机侦察图像的压缩

由于无人机飞行高度高、速度快，所以无人机侦察图像的目标和细节多、数据量大。为了实现图像数据的高速实时传输，需要对无人机侦察图像进行高压缩比压缩；但是在实际应用中，往往只对某些细节和目标感兴趣，要求这些区域清晰可见，这与图像的高压缩比压缩相矛盾，因此，本文提出了一种基于小波的无人机侦察图像压缩方法。该方法将感兴趣的军事目标从原图像中分离开来，对ROI（感兴趣区域）采用低压缩比甚至无损压缩，而对BG（背景区域）采用高压缩比压缩，最后合成两幅图像，解决了无人侦察机图像高压缩比和高质量之间的矛盾。

A Method of ROI Compression of UAV Reconnaissance Image Based on Wavelet

Due to high flight altitude and speed of UAV, the UAV reconnaissance image has many targets and details with a large amount of data. In order to realize live transmission of the image data by high-speed, the UAV reconnaissance image needs to be compressed in high ratio;but in practical applications, people usually feel interested in certain details and targets only, thus these places need to be clearly visible, but this is contradictory to high compression ratio, so this paper puts forward a method of ROI （region of interest） compression of UAV reconnaissance image based on wavelet transfrom. The characteristic of this approach is that the interested military target should be separated from the original image, the ROI is processed with low compression ratio even lossless compression, while other regions are done with high compression ratio, then these two images are combined. Through this way, we can get both high compression ratio and high quality in the UAV reconnaissance image compression.