A Multi-Dimensional Grammar for Modeling Temporal-Spatial Images.
Abstract
Image processing deals generally with well-defined multi-dimensional data with the intention of transforming them into a more useable form. This thesis proposes a grammar which is useful in describing, analyzing, and designing image processing in multi-dimensional terms. The concept of an m-dimensional to n-dimensional dimensioned modeling process (m:ndMP) employed by the proposed grammar allows a systematic approach to modeling image processes. This type of modeling process is designed to be useful in manipulating and relating transformations of multi-dimensional data using networks of directed graphs. Inclusion of a confidence element in the grammar's notation allows inclusion of heuristics in describing image processing algorithms. The structured composition of m:ndMPs parallels that of formal computer programming languages to ease translation of image processing models into software implementations. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1984
- Accession Number
- ADA151945
Entities
People
- B. G. Litherland
Organizations
- Air Force Institute of Technology