EDGE LOCATION AND DATA COMPRESSION FOR DIGITAL IMAGERY
Abstract
A new method for locating edges in digital data with subpixel accuracy and which is invariant to additive and multiplicative changes in the data is presented. In one-dimensional edge patterns it is shown that the edge location is related to the so-called "Christoffel numbers." Also presented is the study of the effect of noise and data compression on edge location. The method is extended to include two-dimensional edge patterns where a line equation is derived to locate an edge with subpixel accuracy. This in turn is compared with the standard Hueckel edge operator. An application of the new edge operator as an edge detector is also provided and is compared with Sobel and Hueckel edge detectors in presence and absence of noise. An adaptive transform coding technique where a new method of assigning regions of each picture to various classes, with unequal number of blocks in each class is presented, to improve the performance in the m.s.e. sense. Also a method of preprocessing the data such that subjective ratings of the reconstructed imagery are improved is presented. Finally, a method is presented to automatically inspect the block boundaries of a reconstructed two-dimensional transform coded image, to locate blocks which are most likely to contain errors, to approximate the size and type of error in the block, and to eliminate this estimated error from the picture. This method uses redundancy in the source data to provide channel error correction. No channel error protection bits or changes to the transmitter are required. This approach can be used when channel errors are unexpected prior to reception.
Degree
Ph.D.
Subject Area
Electrical engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.