Abdesslam Benzinou, Hervé Troadec, Jean Le Bihan, Vincent Rodin, Hélène De Pontual and Jacques Tisseau.
The locally deformable B-Bubble model: An application to growth ring detection on fish otoliths.
SCIA'97, IAPR Scandinavian Conference on Image Analysis, volume 1, pages 181-187, Lappeenranta (Finland), 9-11 June 1997.
Abstract:
One of the major problems encountered during an automatic contour detection is the lack of structure continuity perception. The usual active contour concept is a powerful tool for representing this continuity. However, existing active contour models are not suited for contour detecting in textured noisy and low contrast images. In this paper, we propose a new contour detecting algorithm that we called: the Locally Deformable B-Bubble model (LDBB) based on a parametric B-spline representation. The originality of this approach lies in the basic structure of the model, in the way it operates, and in the expression of a new local driving force. This model evolves by small local deformations, through its control point displacements, proportionally to a local resistance. In addition, it allows the use of high level information that constraint the bubble during its evolution. This model presents a low noise and texture sensitivity. An application to fish otolith images is presented to illustrate model efficiency.
Keywords: Active contour, Local deformation, B-Spline, Otolith, Age and Growth.
[Benzinou97a.pdf]

---