Optimization of Lossless Image Compression Method for Mobile Computing Platform

Model kompresije predikcijom i entropijskim kodiranjem dobar je pristup komprimiranju slikovnih podataka bez gubitka. I korak predikcije i korak entropijskog kodiranja mogu se izvesti na različite načine. Ovisno o odabranoj kombinaciji, u prosjeku, u ovom radu, omjer veličine komprimirane datoteke i originalne slike od oko 60%. Za korištenje na mobilnoj platformi, korak entropijskog kodiranja Vitterovim algoritmom troši mnogo vremena te nije primjenjiv za praktičnu upotrebu. Korak se stoga optimira tako da se umjesto Vitterovog, primijeni algoritam Golomb-Rice koji drastično smanjuje vrijeme izvođenja.Compression model that combines prediction with entropy compression is a good approach to lossless compressing of image data. Both prediction and entropy coding can be implemented in different ways. Depending of their combination, the models that were described here achieve, in average, a 60% ratio between a compressed file and original image. When it is used on a mobile platform, entropy coding via Vitter algorithm uses a significant amount of time, and would not be good for practical uses. Therefore, the model is optimized, replacing Vitter algorithm with Golomb-Rice coding algorithm that drastically reduces execution time