Off-campus UNL users: To download campus access dissertations, please use the following link to log into our proxy server with your NU ID and password. When you are done browsing please remember to return to this page and log out.
Non-UNL users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
Set quantizer
Abstract
Quantizers are widely used in various digital systems. A quantizer is a many-to-one mapping from a possibly infinite set to a finite (generally small) set. A new quantizer, the set quantizer (SQ), is introduced in this dissertation. The SQ was motivated by set expansion of output levels and by the utilization of residual redundancy in the source process. The SQ uses a collection, or set (hence the name), of output points, instead of one output point as in the scalar quantizer, in a quantization region. The SQ maps the input sample to the available output point closest to it while keeping the quantizer output rate the same as the normal scalar quantizer. The selection of the output point in a given set is the key to the success of the SQ. Two schemes are proposed for this purpose: the first borrows ideas from the hidden Markov model and the second is based on a random coding approach. The SQ is tested on memoryless sources, Markov sources and image data. For bit rate greater than two, it is found that the SQ performs better for memoryless sources than all the quantizers found in the literature.
Subject Area
Electrical engineering
Recommended Citation
Bi, Shaolin, "Set quantizer" (1993). ETD collection for University of Nebraska-Lincoln. AAI9415950.
https://digitalcommons.unl.edu/dissertations/AAI9415950