Vertor Quantization and Its Hardware Realization

The fundamental principles of vector quantization (VQ) are introduced and its simulation results presented.VQ coding can reduce the original image's 8bit/pel to the reconstructed image's 0.5 bit/pel, the resolution of the image being 512 by 512. At the same time , hardware realization of the VQ encoder has been implemented. The use of 74 S chips and the replacement of distortion computer by absolute subtracter in the circuit allows the time to finish VQ encoding of image to be reduced a great deal. VQ encoding of a 512 by 512 image can be done within 9 seconds, with a data compres sion ratio of 16:1.