The Prague Stringology Club Workshop '98

Ahmad Daaboul

Local Prediction for Lossless Image Compression

Abstract:
In predictive coding a group of neighboring picture elements is used to select a suitable prediction value for a current pixel. In this paper, we propose two techniques for lossless images compression based on predictive coding. In the first technique which called, the predictors, we replace each pixel in the image by the predicted pixel; we use various schemes to predict the value of a pixel. In the second, which is based on predictor technique, and called optimal prediction schemes, we divide the original image into blocks or lines and seek the best predictor for each (among a selected set of eight) that provides the best prediction. The errors image is encoded through arithmetic coding, during the final step of compression. The gains of compression that we obtained are observed in the lossless image compression.

Download paper: Article in PostScript Article in PDF
 PostScript   PDF