Nested Static Image Coding Algorithm Based On Rate-distortion Optimization Study

With the widespread use of Internet, it has taken several great technical challenges for image compression and transmission. One of better solutions is utilizing characteristic of progressive transmission of embedded still image coding. On severely bandwidth limited channel, it is necessary that images can be rendered with a low resolution as early as possible. The user has the option to terminate the transmission based on the image content or to continue receiving the finer details of the image.In the Internet, users have many different bandwidths and screen resolutions, so image compression algorithms must be suitable for a wide variety of viewing environments and bitstream may be decoded and viewed under many different circumstances. Ideally, the bitsteams with different coding bit-rate that can be gotten by truncating bitsteams at any point of coding and decoding are all optimized , rather than separate optimized for each.The othei circumstances,the speed of coding is quickly enough,and the bitstreams with different bit-rate can be optimized by user's demand.But the bitstreams by trunctioa are not optimized.Both of the circumstances imply that the image coded bitsireams must be embedded,and every one will be dicussed in details in this paper. Connected to low bandwidth channel, the computers with low resolution displays can render an image early in the bitstream without receiving many bits that are only required for high resolution rendering. A user with a high resolution display and a high bandwidth connection to the network should then be able to view a coa-se resolution version of the image which improves as additional bits are read from ihe same bitstream.Under the given distortion conditions, ;he image coding Rate-Distortion theories discuss the max coding bit-rate. Unde the given coding bit-rate,the max coding distortion decrease can be calculated v/ith Rate-Distortion algorithms.The first coding bits have more information and reflect the image outlines,the last coding bits reflect the image details.So the Rate-Distor :ion theory is very important for the embedded still image coding.Embedded still image coding receives great attention recently and becomes one of hot topics in image coripression algorithm. This paper introduces the fundamental theory and kinds of factors that influence the coding quality in details, introduces the main techniques of embedded coding at home and abroad. Based on those introduces, this paper researches embedded still coding methods with Rate-Distortion Optimization .The author thinks, the main obstacles of the development of embedded coding technique in Image Compression are: The speed of encoding and decoding must be quickly enough to be applied in real-time applications. The effect of encoding is so better that the algorithm can render images underany coding bit-rate conditions. Embedded bit streams can be arbitrarily truncated at any point; the coder is not only optimized at the final rate, but also optimized at every trunction point. How can we decide the subband bit allocations and quantization step sizes so that the Mean Squared Errors are minimum?Based on the analysis of the main obstacle of embedded still image coding, this paper focus on the coding order with Rate-Distortion Optimized, subband bit allocations and quantization step sizes. And the main contributions are shown below: This paper presents a kind of embedded still image coding algorithm with coding order with Rate-Distortion Optimized.Based on the steepest Rate-Distortion slope coding order, the bitstream is optimized at arbitrary point. The low coding bit-rate bitstream gotten by truncating high bit-rate one is the same as the bitstream gotten by coding immediately with the coding low bit-rate. The result shows that the speed of algorithm is very quick, the realization is very simple and coding effect is better. This paper analyzes the theory of Rate-Distortion Optimization based on Lagrange Multiplier in details, and presents a kind of Embedded still image co...