### Citations

1885 | Scale-Space and Edge Detection Using Anisotropic Diffusion
- Perona, Malik
- 1990
(Show Context)
Citation Context ... aims to find corner features that persist over image scales, by a sophisticated operation based on the difference of Gaussian operator. By way of example, we show the result of anisotropic diffusion =-=[2]-=- applied to an image of an eye, in comparison with the result by Gaussian filtering, a standard operator. The anisotropic diffusion process in Fig. 1c achieves a more pleasing result than that of Fig.... |

1206 | Active contours without edges
- Chan, Vese
- 2001
(Show Context)
Citation Context ...ge when initialization within a shape can be guaranteed). By way of illustration, we shall use the geometric active contour called the active contour without edges (ACWE), introduced by Chan and Vese =-=[5]-=-, which is the technique that many people compare the result of their own new approach with. Their model uses regional statistics for segmentation, and as such is a region-based level-set model. The o... |

1015 |
Watersheds in digital spaces: an efficient algorithm based on immersion simulations
- Vincent, Soille
- 1991
(Show Context)
Citation Context ...egmentation is floodfill, but this suffers in performance by not including water properties. Instead of model-based methods, some developed the morphological watershed-based region growing techniques =-=[9, 10]-=-. The approach is based on the fact that smooth surfaces can be decomposed into hills and valleys by studying critical points and their gradient. Considering pixel properties (intensity or gradient) a... |

754 | Snakes, shapes, and gradient vector flow
- Chenyang, Prince
- 1998
(Show Context)
Citation Context ...lement modelling and statistical modes from sample data. Thereby a physical model was combined with a statistical model so as to analyse shape variation. The gradient vector flow (GVF) active contour =-=[14]-=- uses a bidirectional external force field that provides long-range capture of object boundaries from either side and is one of the stock approaches for evolving to find arbitrary shapes. There is a r... |

169 | A review of statistical approaches to level set segmentation: integrating color, texture, motion and shape
- Cremers, Rousson, et al.
- 2007
(Show Context)
Citation Context ...volved in rebalancing the level sets used in the formulation, at each iteration of the technique. There have been many developments in the field and newer approaches use texture to guide segmentation =-=[4]-=-. Naturally, performance depends on parameter selection. A major advantage concerns initialization, as the curve evolution techniques evolve to determine a solution within the image which is consisten... |

130 | Diffusion snakes: Introducing statistical shape knowledge into the mumford–shah functional.
- Cremers, Tischhauser, et al.
- 2002
(Show Context)
Citation Context ...Among the most popular modern operators for arbitrary shape extraction are geodesic active contours which evolve to find a target shape contour. Example results by curve evolution are shown in Fig. 2 =-=[3]-=-. These show that the technique can converge to an acceptable result, Fig. 2e—showing the palm of the hand and the fingers—though with an extra finger due to the background, from an initialization whi... |

53 | Force field feature extraction for ear biometrics,”
- Hurley, Nixon, et al.
- 2005
(Show Context)
Citation Context ...approaches. 2. NEW PHYSICAL ANALOGIES FOR FEATURE EXTRACTION 2.1. Force field transform 2.1.1. On deploying the analogy of gravity Hurley et al. developed a transform called the force field transform =-=[15, 16]-=- which uses an analogy to gravitational force wherein mass is equivalent to pixel brightness and distance is measured between pixels. The transform pretends that each pixel exerts a force on its neigh... |

46 | Characterization of neuropathological shape deformations.
- Martin, Pentland, et al.
- 1998
(Show Context)
Citation Context ...feature extraction—in part from the inherent ambiguity of using analogies with search terms. One approach was aimed to separate interesting from uninteresting shape deformations in a class of objects =-=[13]-=-. The approach is general and was demonstrated by separating nonpathological shape variation from pathological deformations in neurological images. To achieve this, a mathematical model was developed ... |

35 | CPM: A deformable model for shape recovery and segmentation based on charged particles
- Jalaba, Wilkinson, et al.
- 2004
(Show Context)
Citation Context ...sed in modern approaches to curve evolution. Therefore, the magnetic force can be easily combined with level sets. Electrostatic field theory has also been adopted in image segmentation. Jalba et al. =-=[8]-=- presented a charged particle model that simulates the particle movements in an electrostatic field. A set of positively charged free particles is positioned in the image analogical field distributed ... |

33 | Force field energy functionals for image feature extraction,
- Hurley, Nixon, et al.
- 2002
(Show Context)
Citation Context ...approaches. 2. NEW PHYSICAL ANALOGIES FOR FEATURE EXTRACTION 2.1. Force field transform 2.1.1. On deploying the analogy of gravity Hurley et al. developed a transform called the force field transform =-=[15, 16]-=- which uses an analogy to gravitational force wherein mass is equivalent to pixel brightness and distance is measured between pixels. The transform pretends that each pixel exerts a force on its neigh... |

30 |
Watershed-based Segmentation and Region Merging,”
- Bleau, Leon
- 2000
(Show Context)
Citation Context ...egmentation is floodfill, but this suffers in performance by not including water properties. Instead of model-based methods, some developed the morphological watershed-based region growing techniques =-=[9, 10]-=-. The approach is based on the fact that smooth surfaces can be decomposed into hills and valleys by studying critical points and their gradient. Considering pixel properties (intensity or gradient) a... |

13 |
Anisotropic diffusion pyramids for image segmentation,” in
- Acton, Bovik, et al.
- 1994
(Show Context)
Citation Context ...ds irregular boundaries, over-segmentation and small holes. One of the most popular uses of heat in image processing is anisotropic diffusion which is a state-of-art process for image enhancement. In =-=[11]-=-, an anisotropic diffusion pyramid was introduced for region-based segmentation. The pyramid is constructed using the scale-space representation of the anisotropic diffusion. In [12], the anti-geometr... |

10 | Magnetostatic Field for the Active Contour Model: A Study in Convergence
- Xie, Mirmehdi
- 2006
(Show Context)
Citation Context ...age processing Though there is no other work similar to our approach using a gravitational analogy, there are other force-based approaches motivating image processing feature extraction operators. In =-=[6, 7]-=-, magnetostatic theory was combined with an active contour model. For a moving charge inside a magnetic field, there are magnetic forces introduced by the field acting on it. Further, since currents a... |

8 | Anti-geometric diffusion for adaptive thresholding and fast segmentation
- Manay, Yezzi
- 2003
(Show Context)
Citation Context ...e enhancement. In [11], an anisotropic diffusion pyramid was introduced for region-based segmentation. The pyramid is constructed using the scale-space representation of the anisotropic diffusion. In =-=[12]-=-, the anti-geometric heat-flow model was introduced for the segmentation of regions. Here, anti-geometric heat flow is represented as diffusion The Computer Journal, Vol. 54 No. 1, 2011sat Serials Rec... |

6 | Correction to “MAC: Magnetostatic active contour model”
- Xie, Mirmehdi
(Show Context)
Citation Context ...age processing Though there is no other work similar to our approach using a gravitational analogy, there are other force-based approaches motivating image processing feature extraction operators. In =-=[6, 7]-=-, magnetostatic theory was combined with an active contour model. For a moving charge inside a magnetic field, there are magnetic forces introduced by the field acting on it. Further, since currents a... |

6 | Water Flow Based Complex Feature Extraction
- Liu, Nixon
- 2006
(Show Context)
Citation Context ...tion point, we need to determine where water will flow, according to selected image forces. The basis of our approach uses pressure differential equation (8) moderated by surface tension and adhesion =-=[17, 18]-=-. The water element can move outwards in any direction for which the component of velocity is positive. However, only if the velocity in the direction is sufficiently large, can the element break thro... |

5 | On Using Physical Analogies for Feature and Shape Extraction in Computer Vision
- Nixon, Liu, et al.
- 2008
(Show Context)
Citation Context ... heat. We shall outline the basis and results of our new approaches using physical analogies, and further detail can be found in earlier descriptions of these works [15–20]. Beyond an earlier version =-=[21]-=- this paper has an extended description of each of the techniques, in particular showing new feature extraction by gravitational analysis and how low-level feature extraction can be achieved by using ... |

5 |
Shape Classification using Multiscale Fourier-Based Description in 2-D Space
- Direkoğlu, Nixon
- 2008
(Show Context)
Citation Context ...ies better than ACWE especially at the middle and at the bottom of the lung image. We have also extended the heat concept for feature extraction in images, via a multiscale Fourier-based descriptions =-=[23]-=- (note that the original presentation by Fourier concerned heat). 3. CONCLUSIONS AND FURTHER WORK New approaches to shape extraction and to low-level processing can be achieved by using physical analo... |

4 |
Image and Volume Segmentation by Water Flow
- Liu, Nixon
- 2007
(Show Context)
Citation Context ...tion point, we need to determine where water will flow, according to selected image forces. The basis of our approach uses pressure differential equation (8) moderated by surface tension and adhesion =-=[17, 18]-=-. The water element can move outwards in any direction for which the component of velocity is positive. However, only if the velocity in the direction is sufficiently large, can the element break thro... |

3 | Shape Extraction via Heat FlowAnalogy - Direkoğlu, Nixon - 2007 |

2 |
Low Level MovingFeature Extraction Via Heat Flow Analogy
- Direkoğlu, Nixon
- 2006
(Show Context)
Citation Context ...ons of the fitted model in upper and lower slices are also shown in Fig. 12. 2.3. Using heat for feature extraction 2.3.1. Low-level features It is possible to use heat flow to determine moving edges =-=[19]-=-. The algorithm combines anisotropic and isotropic heat flow. To filter noise anisotropic diffusion is first applied separately to each image of a walking subject Fig. 13a–c, here to Framei+1, Framei,... |

1 |
X-ray image processing for high voltage cable inspection
- Robinson, Lewin, et al.
- 2005
(Show Context)
Citation Context ...e image on the right shows the evolution which led to determining these positions. The gravitational approach, apart from its use in Biometrics, has been also applied to high-voltage cable inspection =-=[22]-=- wherein the difference between applying gravitational field analysis to a sample and to a uniform image highlighted defects in the material. 2.1.2. Convergence feature extraction A further advantage ... |