Publication View

Scaling Results for the Variational Approach to Edge Detection (2007)

Abstract
In another paper the first author presented an asymptotic result concerning the Mumford-Shah functional. In this paper an algorithm is presented which implements the scaling suggested by the asymptotic results to obtain accurate localization of edges even when only large scale edges are being detected. An approximation to E is also considered and the implications of the asymptotic results to this functional are also examined.. Prepared in cooperation with AT&T Bell Labs, Murray Hill, NJ. Sponsored in part by Army Research Office.

Publication details
Download http://handle.dtic.mil/100.2/ADA459531
Contributors MASSACHUSETTS INST OF TECH CAMBRIDGE LAB FOR INFORMATION AND DECISION SYSTEMS
Repository Defense Technical Information Center OAI-PMH Repository (United States)
Keywords NUMERICAL MATHEMATICS, OPERATIONS RESEARCH, CYBERNETICS, *MATHEMATICAL MODELS, *ALGORITHMS, *IMAGE PROCESSING, *EDGES, OPTIMIZATION, FUNCTIONAL ANALYSIS, COMPUTER VISION, VARIATIONAL PRINCIPLES, *EDGE DETECTION, *VARIATIONAL MODELS, *ASYMPTOTIC THEOREMS, SEGMENTATION
Language eng