Skip to Main Content
The SAR image data must be compressed efficiently so that the requirements for transmission bandwidth and storage space, which are brought by large amount of data on SAR image, can be reduced. The traditional methods of SAR image compression based on wavelet transformation can only decompose low frequency sub-bands, resulting in the loss of important information of high frequency sub-bands. Aiming at the problem mentioned above, in this paper, a self-adaptive SAR image compression algorithm based on wavelet transformation is presented. Firstly, the speckle noise is reduced through wavelet soft threshold. And then the importance of sub-bands is judged according to the energy indicators, to further decompose the more important ones. After comleting the decomposition, all the sub-bands are quantized by the minimum error rule at the same bit rate to achieve the self-adaptive SAR image compression. It is shown that the algorithm can well protect the high frequency details of SAR image and achieve better compression results as well.