Limits...
Convolutional virtual electric field for image segmentation using active contours.

Wang Y, Zhu C, Zhang J, Jian Y - PLoS ONE (2014)

Bottom Line: Gradient vector flow (GVF) is an effective external force for active contours; however, it suffers from heavy computation load.Meanwhile, the CONVEF model can also be implemented in real-time by using FFT.Experimental results illustrate these advantages of the CONVEF model on both synthetic and natural images.

View Article: PubMed Central - PubMed

Affiliation: School of Computer Science, Tianjin University of Technology, Tianjin, China.

ABSTRACT
Gradient vector flow (GVF) is an effective external force for active contours; however, it suffers from heavy computation load. The virtual electric field (VEF) model, which can be implemented in real time using fast Fourier transform (FFT), has been proposed later as a remedy for the GVF model. In this work, we present an extension of the VEF model, which is referred to as CONvolutional Virtual Electric Field, CONVEF for short. This proposed CONVEF model takes the VEF model as a convolution operation and employs a modified distance in the convolution kernel. The CONVEF model is also closely related to the vector field convolution (VFC) model. Compared with the GVF, VEF and VFC models, the CONVEF model possesses not only some desirable properties of these models, such as enlarged capture range, u-shape concavity convergence, subject contour convergence and initialization insensitivity, but also some other interesting properties such as G-shape concavity convergence, neighboring objects separation, and noise suppression and simultaneously weak edge preserving. Meanwhile, the CONVEF model can also be implemented in real-time by using FFT. Experimental results illustrate these advantages of the CONVEF model on both synthetic and natural images.

Show MeSH

Related in: MedlinePlus

In each panel, from left to right, (a) synthetic image of two objects and its edge map; convergence and the close-up of force field of (b) VEF snake, (c) CONVEF snake with n = 3.0, h = 0.0, (d) VFC snake using m2 with .
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4216009&req=5

pone-0110032-g007: In each panel, from left to right, (a) synthetic image of two objects and its edge map; convergence and the close-up of force field of (b) VEF snake, (c) CONVEF snake with n = 3.0, h = 0.0, (d) VFC snake using m2 with .

Mentions: Fig. 5 demonstrates the use of large n for edge preserving when the CONVEF snake is employed to locate objects. One can also employ large n to separate two closely neighboring objects, especially when one edge is weak and the other is strong. In fact, to separate objects is essential to preserve the edge of each object. We demonstrate this particular application using a synthetic image. Fig. 7(a) shows the original image, where there are one gray disk and one white rectangle on the black background and there are just three pixels between two objects. The edge of the disk is weak and that of the rectangle is strong. Fig. 7 (b) shows that the VEF snake moves across the weak edge and sticks to the strong one. Fig. 7 (c) shows the result of the CONVEF snake with and . It is clear that the CONVEF snake correctly tells apart these two closely neighboring objects. Fig. 7(d) shows the result of VFC snake using with . Since the VFC snake prefers large ζ to enlarge the capture range, but ζ is just 3 in this example and the capture range is very low, so the initial contour is close to the disk. However, the snake contour still leaks out to the rectangle.


Convolutional virtual electric field for image segmentation using active contours.

Wang Y, Zhu C, Zhang J, Jian Y - PLoS ONE (2014)

In each panel, from left to right, (a) synthetic image of two objects and its edge map; convergence and the close-up of force field of (b) VEF snake, (c) CONVEF snake with n = 3.0, h = 0.0, (d) VFC snake using m2 with .
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4216009&req=5

pone-0110032-g007: In each panel, from left to right, (a) synthetic image of two objects and its edge map; convergence and the close-up of force field of (b) VEF snake, (c) CONVEF snake with n = 3.0, h = 0.0, (d) VFC snake using m2 with .
Mentions: Fig. 5 demonstrates the use of large n for edge preserving when the CONVEF snake is employed to locate objects. One can also employ large n to separate two closely neighboring objects, especially when one edge is weak and the other is strong. In fact, to separate objects is essential to preserve the edge of each object. We demonstrate this particular application using a synthetic image. Fig. 7(a) shows the original image, where there are one gray disk and one white rectangle on the black background and there are just three pixels between two objects. The edge of the disk is weak and that of the rectangle is strong. Fig. 7 (b) shows that the VEF snake moves across the weak edge and sticks to the strong one. Fig. 7 (c) shows the result of the CONVEF snake with and . It is clear that the CONVEF snake correctly tells apart these two closely neighboring objects. Fig. 7(d) shows the result of VFC snake using with . Since the VFC snake prefers large ζ to enlarge the capture range, but ζ is just 3 in this example and the capture range is very low, so the initial contour is close to the disk. However, the snake contour still leaks out to the rectangle.

Bottom Line: Gradient vector flow (GVF) is an effective external force for active contours; however, it suffers from heavy computation load.Meanwhile, the CONVEF model can also be implemented in real-time by using FFT.Experimental results illustrate these advantages of the CONVEF model on both synthetic and natural images.

View Article: PubMed Central - PubMed

Affiliation: School of Computer Science, Tianjin University of Technology, Tianjin, China.

ABSTRACT
Gradient vector flow (GVF) is an effective external force for active contours; however, it suffers from heavy computation load. The virtual electric field (VEF) model, which can be implemented in real time using fast Fourier transform (FFT), has been proposed later as a remedy for the GVF model. In this work, we present an extension of the VEF model, which is referred to as CONvolutional Virtual Electric Field, CONVEF for short. This proposed CONVEF model takes the VEF model as a convolution operation and employs a modified distance in the convolution kernel. The CONVEF model is also closely related to the vector field convolution (VFC) model. Compared with the GVF, VEF and VFC models, the CONVEF model possesses not only some desirable properties of these models, such as enlarged capture range, u-shape concavity convergence, subject contour convergence and initialization insensitivity, but also some other interesting properties such as G-shape concavity convergence, neighboring objects separation, and noise suppression and simultaneously weak edge preserving. Meanwhile, the CONVEF model can also be implemented in real-time by using FFT. Experimental results illustrate these advantages of the CONVEF model on both synthetic and natural images.

Show MeSH
Related in: MedlinePlus