Intelligent Vision Systems,
Abstract
Research on intelligent systems for image understanding focusses on a successor to the ACRONYM system, and its application in a rule-based stereo mapping and interpretation system. Some elements of rule-based stereo system have been implemented. A new modeling system is under construction, and a new graphics system for display of generalized cylinders has achieved initial results. Research has continued on segmentation/aggregation in the figure-ground problem for grouping candidate objects. Implementation experiments are underway for an array of vision processors. Fundamental mathematical results have been obtained on matching processes. Inference rules for interpreting surfaces from images were demonstrated formally in a mathematical logic programming system. Results have been obtained in specializing certain vision programs by automatic methods to produce efficient programs. The objectives of this research are to develop algorithms for high performance image understanding modules, to implement an intelligent vision system, and to demonstrate its application in photointerpretation and cartography. The ACRONYM system was developed as the first intelligent system. Research has shifted to its SUCCESSOR.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 1983
- Accession Number
- ADP001191
Entities
People
- Thomas O. Binford
Organizations
- Stanford University