Paper
29 March 2016 Iterative weighted average diffusion as a novel external force in the active contour model
Author Affiliations +
Abstract
The active contour model has good performance in boundary extraction for medical images; particularly, Gradient Vector Flow (GVF) active contour model shows good performance at concavity convergence and insensitivity to initialization, yet it is susceptible to edge leaking, deep and narrow concavities, and has some issues handling noisy images. This paper proposes a novel external force, called Iterative Weighted Average Diffusion (IWAD), which used in tandem with parametric active contours, provides superior performance in images with high values of concavity. The image gradient is first turned into an edge image, smoothed, and modified with enhanced corner detection, then the IWAD algorithm diffuses the force at a given pixel based on its 3x3 pixel neighborhood. A forgetting factor, φ, is employed to ensure that forces being spread away from the boundary of the image will attenuate. The experimental results show better behavior in high curvature regions, faster convergence, and less edge leaking than GVF when both are compared to expert manual segmentation of the images.
© (2016) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Ilya S. Mirov and Arie Nakhmani "Iterative weighted average diffusion as a novel external force in the active contour model", Proc. SPIE 9788, Medical Imaging 2016: Biomedical Applications in Molecular, Structural, and Functional Imaging, 97881M (29 March 2016); https://doi.org/10.1117/12.2217168
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Image segmentation

Diffusion

Convolution

Error analysis

Image processing algorithms and systems

Tumors

Machine vision

Back to Top