Optimum quantization.
Year: 1965
Type: report
Abstract: The exact expression for the quantization error as a function of the parameters defining the quantizer, the error-weighting function, and the amplitude probability density of the quantizer-input signal is presented.An algorithm is developed that permits us to determine the specific values of the quantizer parameters that define the optimum quantizer.This algorithm is then extended so that optimum quantizers can be determined for the case in which the quantizer-input signal is a message signal contaminated by noise.In each of these cases the algorithm is based on a modified form of dynamic programming and is valid for both convex and nonconvex error-weighting functions.Examples of optimum quantizers designed with the first of these two algorithms for a representative speech sample are presented.The performance of these optimum quantizers is compared with that of the uniform quantizers.
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7051544267
