A new Algorithm for Lossless Compression applied to two-dimensional Static Images

In this paper, a new lossless compression algorithm method for two-dimensional continuous-tone images is presented. The proposed method achieves higher compression ratios than universal methods. Traditional methods use coding techniques based on reduction of the existing redundancy in the data or pixels, mainly applying methods based on statistical models, substitution models, dictionary models, predictive models, wavelet transform domain, etc... The proposed method is based on a model which incorporates a new algorithm (INA Algorithm). This method consists of three steps: segmentation, compression algorithm and encoding. In the first, the whole image is continuously divided into blocks of a fixed size. Each block generates two separate groups: a sequence of the different pixels of the block sorted in increasing order and the other one in a sequence formed of pixels organised in groups of ordered pairs. The second one is the new development and main contribution of this method. A new algorithm processes in parallel the execution of the two sequences obtained in the segmentation process by means of a data structure in the form of a binary tree and a predictive scheme respectively. Lastly, a fast and efficient encoder is used to generate the binary codes. The parallel execution of this algorithm reduces response times of the compression and decompression cycles. This method ensures a compression ratio of 1:2 to 1:4 with very high probability. In order to evaluate the advantages of this method compared with the conventional lossless methods and the JPEG-LS standard, experimental studies have been carried out using a set of ISO standard images. From the experimental results we can asserted that the proposed method obtains a higher compression ratio. Key-Words: Lossless compression, image compression, binary tree