Research On Line-based Low Memory Wavelet Image Compression

In case of non-compression, image measure very huge by the digital data of signal, which bring many difficulty and problems for the transform, storage and disposal of image information. So the compression of the image data is very essential. The compression of digital images has been a topic of research for many years and a number of image compression standards have been created for different applications. The new image compression algorithms are emerging constantly too. The wavelet is viewed as the strong tool to analyze non-stable signal, has gotten relatively extensive application in the image compression. In going on compression based on wavelet transform, all the wavelet coefficients need to be buffered at the course of transform. The memory is taken up relatively bigly, and become the bottlenecks that restrain the wavelet image compression development gradually. There are some effectively researches of the memory usage in wavelet coding recently. Among these methods, some concentrate on the DWT, some only consider compressing algorithms, or only consider the encoder, decoder one of the two.-This paper adopt the line-based wavelet transform, aiming at the whole systematic memory, propose a method to reduce the memory at both encoding and decoding. This line-based wavelet transform is essentially different with the traditional method at the orders of filter. The images are read line by line and only the minimum required number of lines is kept in memory. A completely separable implementation would require that all the lines be filtered before column filtering starts and thus memory sizes of the order of the image size will be require. The approach of this paper is to start column filtering as soon as a sufficient number of lines, as determined by the filter length, have been horizontally filtered. This will allow us to store in memory only a reduced number of input lines. With this approach the memory needs of the encoder and decoder depend only on the width of the image (rather than the total size as in a traditional row column filtering implementation) with significantly enhances the scalability of the system.The main contributions of our work are two. First, we introduce a line-based approach for the implementation of the wavelet transform. Beginning analyze with one-dimensional wavelet transform memory utilization, divide the necessary memory of the system into the filter memory and synchronization memory. At last we present the flow chat for total system. Second, we propose a novel context-based encoder,which requires no global information and stores only a local set of wavelet coefficients. Thus realize the low memory of the code scheme. With respect to existing coders (SPIHT, JPEG2000), the degradation in performance is modest. In terms of memory utilization, our results show reductions of almost two orders of magnitude with respect to widely available implementations of wavelet image coders.