L.F. Costa and A.C.P. Veiga (Brazil)
Genetic Algorithms (GA), Quantization Table (Q-Table), and Natural Selection.
In several situations, images are processed in JPEG standard with a Quantization Table without any knowledge of where does the coefficient come from, and there is little literature to produce a robust quantization table. The fidelity loss in JPEG coding occurs entirely in quantization [1], and it becomes very important to the compression/fidelity performance of JPEG coding. The proposal of this article is to generate a Quantization Table that is more robust for medical images using Genetic Algorithms method where the Signal Noise Relation and Median Quadratic Error are observed in each image as a parameter of the quality of the reconstructed image.
Important Links:
Go Back
