F. C. Cesbron, F. J. Malassenet. Comparison of mean-square and absolute value distortion measures in fractal coding of still images. In Proceedings of the VIIIth European Signal Processing Conference EUSIPCO'96, Trieste, 1996.
Fractal coding often blurs or smoothes images. In particular, edges are poorly coded due to the choice of the distortion measure. Indeed, mean square error possesses so edge preserving property. The solution proposed in this article is to code using fractals and to introduce the L1 norm or the absolute value distortion measure. The visual quality of the coded images with the L1 norm is improves. However, the computational complexity of the coder is drastically increased. It may be kept low by reprocessing using an edge detection scheme to select the pertinent measure. The reconstruction algorithm remains the same.
@InProceedings{CeMa96,
Author = {Cesbron, F. C. and Malassenet, F. J.},
Title = {Comparison of mean-square and absolute value distortion measures in fractal coding of still images},
BookTitle = {Proceedings of the VIIIth European Signal Processing Conference EUSIPCO'96},
Address = {Trieste},
Month = {},
Year = {1996}
}