Title :
An optimization-based mesh-generation method for image representation
Author :
Seyedali Mostafavian;Michael D. Adams
Author_Institution :
Dept. of Electrical and Computer Engineering, University of Victoria, BC, V8P 5C2, Canada
Abstract :
In earlier work, Tu and Adams proposed the ERDED method, an effective technique for generating triangle-mesh models of digital images through explicit representation of discontinuities. In this paper, we propose a modified version of the ERDED method that generates a triangle mesh which better approximates the original image using an optimization-based algorithm. The proposed method is shown to produce image approximations of higher quality than those obtained by the ERDED method, both in terms of squared error and subjective quality. Moreover, this improvement in quality comes at a relatively modest computational cost, with the proposed method typically taking only a few seconds of computation time.
Keywords :
"Image edge detection","Wheels","Optimization","Computational modeling","Adaptation models","Lead","Analytical models"
Conference_Titel :
Communications, Computers and Signal Processing (PACRIM), 2015 IEEE Pacific Rim Conference on
Electronic_ISBN :
2154-5952
DOI :
10.1109/PACRIM.2015.7334840