A novel predictor function for lossless image compression

These days we encounter high volume of information around us, to store and reuse theses information they need to be saved. Images also need these saved information. In normal circumstances the volume is extremely high, for example to save an image we may need a thousand bits, Therefore in the world of new technology with the variety of media storage with high capacity, the need for a way to decrease volume of data for a good quality images is felt more than ever. One of many methods that presented so far is wonderful predictive method. In this paper, a novel predictor function for lossless image compression is presented. While almost all existing predictor functions are imperfect, we try to discover the novel function with high generalization and best Performance with two main aims. First, the exact and correct prediction for gray-scale value of pixels. Second, preserve and transfer the correct change-over of gray-scale value. We consider the MED algorithm for the default and base method and compare with our function with implementation on some instances.