Study on Applications of Evolutional Hardware in Lossy Image Compression
Song Jian-zhong · Journal of Electron Devices · 2004
A structure of evolvable hardware applied to image compression is put forward, and the simulation results fully verify that its effect is better than the previous. In the process of the image compression, the use of evolution - function makes this evolvable hardware be able to search for the best prediction function. Because the evolution unit and the best - function unit are parallel while they are working, and most of the genetic operation is completed by hardware, the speed of compression is very high. The simulation results show that the compression ratio is better than JPEG, Fractal, Wavelet and maximum can attain 201:1 when PSNR = 24-30 dB.