Abstract: In the domain of surveying and mapping , users present rigorous requirements for the image quality. But there are noises in the image data, so lossy compression is feasible. In this paper, a method of remote image lossy compression in control is presented. First, the feasibility of remote image lossy compression is analysed. Then the method of two layer and two dimension wavelet transform to decompose original image is adopted, and 7 high frequency quantizers according to remote image's feature is created. At last, the relation between the compresion ratio and peak peak SNR at the different high frequency quantizers is stakied.