Uav Reconnaissance Image Compression Based On Integer Wavelet

The advance of digital image technique produced a great deal of UAV reconnaissance images data, which caused many problems for storage and transmission. Because reconnaissance image has weak local correlation, great data quantity, and abundance detail, traditional compression methods have some limits in some degree. In recent decade, however, wavelet transform has been studied and applied in image compression extensively because of its good time-frequency features, excessive differentiate features and tally with human visual system.This thesis first discussed the essentiality and feasibility of iamge compression, introduce the method of image comptession and the metewand of image compression performance. Sencondly, the theory of wavelet transform are studied. Thirdly through analyzing zhe shortcomings of the traditional wavelet, the lifting scheme using to construct the second generation wavelet is inducted. We give the algorithm of getting the coefficients of new lifting wavelet and the procedure of realizing the traditional wavelet with lifting scheme, and introduce the 9/7 integer wavelet transform used to the lossless compression. Lastly, base on the good and bad of SPIHT (Set Partitioning In Hierarchiacl Trees), the thesis puts forward the improved omnibus coder arithmetic based on differential pulse code modulation (DPCM) and IWT. At first we used the compression result as the standard, and the spy image as the standard training image. The compression ratio, PSNR and reserved energy percentage after compression are taken as parameters to look for the best wavelet filter. Secondly reconnaissance image is used to code by the DPCM, to be has the difference image. Then the different image is decomposed by using 9-7 tap integer lifting wavelet. Finally, the SPIHT are used to code all the sub-band after transformed. Adopting MATLAB to programme this thesis, from the PSNR, CPU time-consuming, compression ratio and visual parameters such as the human eye under the right mix of programs to improve the simulation to achieve, the result simulated indicates:Improved hybrid algorithm for this article in the same bit-rate PSNR than SPIHT algorithm, especially for low bit rate in case of image restoration quality significantly improved.