Sar Image Data Compression Techniques

The image compression technique is how to move redundant information which is independent of image quality and denote the image with least bits under the premise of ensuring image quality. Synthetic aperture radar with the advantages of high-resolution, full time and all-weather is used widely in military and civil application. SAR image compression is a very popular research domian.SAR images have different characteristics compared with ordinary optical images for its coherent trait. A new compression method for reducing speckle noise in SAR image based on optimal wavelet packet by analyzing the characteristics of SAR image data is proposed in this thesis, combined with multi-resolution analysis of wavelet packet transform. First, the significance, developmental status, application of SAR image compression in domestic and overseas are introduced, the traditional image compression coding technologys are also expatiated. Second, the statistical characteristics of SAR images are focused on, wavelet transform theory, Mallat algorithm and lifting scheme of image are introduced. Third, aiming at SAR image speckle noise, a wavelet-domain circulating threshold denoising algorithm is proposed and better denoising effect is obtained; for rich texture details of SAR image , the principle of wavelet packet decomposition and the criteria of best wavelet packet tree selection are alse analysized, and the wavelet coefficients coding algorithm based on ESOT is studyed, SAR image compression algorithm for reducing speckle noise based on best wavelet packet is proposed and excellent compression results are achieved. In addition, the wavelet decomposition level and wavelet basis on compression performance of SAR image are also analyzed, and a lot of important conclusions are getted. At last, the work, and present deficiency and research orientation in future are summarized.A number of optical images and SAR images are selected and a large number of simulation experiments are made in this thesis. By comparing and analyzing simulation results, some conclusions which have important referenced value and practical application value are got. There are two main innovation in this thesis: one is proposing an optimization threshold denoising algorithm: a circulating algorithm for SAR Image Based on the Wavelet Threshold Denoising Method; the other is putting forward an image compression algorithm based on best wavelet packet, combining speckle suppression. A large number of examples show that their impact on improving the performance of SAR image compression and the innovations have a certain degree of application value.