Lossless image compression based on data folding
The paper presents an approach for lossless image compression in spatial domain for continuous-tone images using a novel concept of image folding. The proposed method uses the property of adjacent neighbor redundancy for prediction. In this method, column folding followed by row folding is applied iteratively on the image till the image size reduces to a smaller pre-defined value. For column folding, elements of even columns are subtracted from elements of odd columns. Thereafter, row folding is applied on odd columns in a similar fashion. In row folding, even rows are subtracted from odd rows and the resultant odd rows are used for next iteration. The difference data, thus obtained, is stored in a tile format; which along with the reduced image is encoded and transmitted. The proposed method is compared with the existing standard lossless image compression algorithms and the results show comparative performance. Data folding technique is a simple approach for compression that provides good compression efficiency and has lower computational complexity as compared to the standard SPIHT technique for lossless compression.