Accession Number : ADA332074

Title :   Applications of Geometric Invariants to Computer Vision - AASERT Grant.

Descriptive Note : Final rept. 1 Jul 94-30 Sep 97,

Corporate Author : NEBRASKA UNIV LINCOLN

Personal Author(s) : Bhattacharya, Prabir

PDF Url : ADA332074

Report Date : 12 AUG 1997

Pagination or Media Count : 7

Abstract : This is an AASERT grant that is a supplementary award to the parent AFOSR Grant FR962O-94-1-OO2. The objective was to train a graduate student who must be a U.S. citizen. We have achieved this objective and the student will be awarded a Ph.D. in Computer Science in August 1997. Also, the student published several papers in refereed journals and gave oral presentations in two international conferences. We developed a method for object recognition by extracting conics from digitized images and computing their invariants. The proposed method was extensively tested under noisy conditions and the results were highly satisfactory. We also investigated the geometrical properties of the Legendre transform that are of interest in image understanding applications. Finally, we developed a matching algorithm for space curves, which reduces the problem to finding the eigenvalues of a 4-dimensional matrix.

Descriptors :   *MANAGEMENT INFORMATION SYSTEMS, *STUDENTS, *AWARDS, SYMPOSIA, DOCUMENTS, INTERNATIONAL, COMPUTER VISION, GRADUATES, GRANTS.

Subject Categories : Economics and Cost Analysis
      Personnel Management and Labor Relations
      Computer Systems Management and Standards

Distribution Statement : APPROVED FOR PUBLIC RELEASE