Rate Control Algorithm Of The Satellite Images In Transmission

With the improvement of the resolution of the satellite images, the data that need to be downloaded from the satellite will enhance constantly. However, the satellite downlink channel bandwidth is limited. Undoubtedly, it needs to complete efficient image compression coding tasks on the satellite. JPEG2000has several outstanding compression characteristics, such as, high compression efficiency, good error resilient ability, code stream progressive transmission and so on, which is getting more and more widely used. As for the satellite images, based on wavelet transform JPEG2000algorithm can be regarded as a kind of ideal satellite image compression scheme. However, the pros and cons of JPEG2000rate control algorithm will affect the entire image coding efficiency and compresse the image quality directly. Therefore, research the more effective JPEG2000rate control algorithm to reduce the coding computing redundancy, improve the encoding speed and reduce the resources consumption of memory occupancy, which have an important practical significance to satellite image transmission.Because the optimized truncation embedded block coding (EBCOT) of JPEG2000algorithm has very high complexity, which influences its application. EBCOT is consisted of two parts: Tier1and Tier2. Tier1code contains a larger portion of computation time. The post compression rate distortion (PCRD) strategy is proposed by Tier2, which discards redundant coding passes. However, PCRD algorithm need code all of passes in Tier1, and need store all the encoded coding passes. At a low bit rate, some of the coding passes are not included in the final output. According to the problems of JPEG2000PCRD in Tier1, some improved rate control algorithm:(1) According to the first threshold that the minimum rate distortion slope of the first code and circulation judge makes the code passes that the rate distortion slopes are greater than the threshold in each coding block as "1" state, and accumulate rate at the same time. After encoding all of the code blocks, if the accumulated rate is less than the target rate, decreasing the threshold and continue to the process until the accumulated rate exceeds the target rate. No need code the code passes for less than the threshold ("0" state), so a large redundant code passes will be skipped.(2) According to the downward trend roughly of the number of bit planes required for the code block and the three properties of the rate distortion slope to achieve the rate distortion slope as the threshold of tile1encoding, and the code passes that less than this threshold are skipped, so it does not need to code all code passes.(3) Making the number of bit planes required for the code block from bigness to smallness by a simple formula to achieve the rate distortion slope as the threshold of tile1encoding, and truncating needn’t the code passes in advance, so improves the coding efficiency.(4) First of all, making code block to sort by the number of bit planes required for the code block, and then use the characteristics between the compression ratio and the target code rate to propose an adaptive Formula, through the adaptive Formula to calculate the near optimal rate distortion slope, truncate a lot don’t need coding pass in advance.Experimental result shows that the proposed algorithm can be made to reduce the Tier1redundancy and reduce the Tier2search optimal rate-distortion slope threshold value and the range of optimal truncation by keeping the original peak signal to noise ratio (PSNR), which reduces the computation, memory of the satellite image and improve the coding efficiency.