Two-dimensional sequential algorithms for image restoration and enhancement
Abstract
The development of a mathematical model for restoration and enhancement of noisy images is described. Two spatially invariant two-dimensional Markov models are developed assuming separable, exponential, diamond symmetric and circularly symmetric image correlation functions. A sequential minimum variance estimation algorithm for a discrete two-dimensional image field modeled by a nonsymmetric half-plane model is derived. A modification of the algorithm to facilitate in-place computation is presented and its implementation discussed. Approximation techniques to reduce the processing time and memory requirements are presented.
- Publication:
-
Ph.D. Thesis
- Pub Date:
- December 1979
- Bibcode:
- 1979PhDT........52P
- Keywords:
-
- Algorithms;
- Image Enhancement;
- Image Processing;
- Mathematical Models;
- Sequencing;
- Approximation;
- Computer Storage Devices;
- Image Correlators;
- Markov Processes;
- Signal To Noise Ratios;
- Two Dimensional Models;
- Electronics and Electrical Engineering