UMP Institutional Repository

An energy functional model by gradient vector-driven active contour for local fitted image segmentation

Wang, Jing and Hai, Tao and M. Nomani, Kabir (2019) An energy functional model by gradient vector-driven active contour for local fitted image segmentation. In: BCPT, Basic & Clinical Pharmacology & Toxicology: 2019 Asia Pacific Conference on Medical and Health Science, 1-3 June 2019 , Seoul, South Korea. p. 218., 125 (S1).

[img]
Preview
Pdf
An energy functional model by gradient.pdf

Download (56kB) | Preview

Abstract

To design a gradient vector-driven active contour local fitted image segmentation model based on information entropy for analyzing the construction of the active contour model by the variational and level set methods for validating the proposed model theoretically and simulation experiments. Firstly, several important active contour models based on image boundary features are introduced, and the existing problems are analyzed in depth, and the causes of the problems are pointed out. Next, the non-conservative behavior of the gradient vector flow field is studied in depth, and an important conclusion about the flow field divergence of the gradient vector is obtained in the local fitted image segmentation model. On this basis, a new energy functional is constructed to measure the flux of the gradient vector flow field through the active curve, and transform the image segmentation problem into the minimum value of the energy functional. Finally, a new active contour model is constructed using the gradient flow of the above energy functional.

Item Type: Conference or Workshop Item (Paper)
Uncontrolled Keywords: gradient vector-driven active contour; local fitted image segmentation
Subjects: Q Science > QA Mathematics > QA76 Computer software
Faculty/Division: Faculty of Computing
Depositing User: Dr. Muhammad Nomani Kabir
Date Deposited: 23 Dec 2019 01:09
Last Modified: 23 Dec 2019 01:09
URI: http://umpir.ump.edu.my/id/eprint/25829
Download Statistic: View Download Statistics

Actions (login required)

View Item View Item