M. Li, C. Kambhamettu, and M. Stone (USA)
Image Processing, Snakes, Tracking, Edge Extraction.
A novel snake model suitable for edge extraction of band shape objects is presented in this paper. Based on the proposed model, an edge tracking system, EdgeTrak, has been developed which is being used by speech scientists in speech research and other related applications. Unlike the classical active contour models which only use gradi ent of the image as the image force, the proposed snake model incorporates the edge gradient and intensity infor mation in specific regions around each snake element. It can be used to extract edges that are open or closed con tours, which makes it different from other active contour models that use homogeneity of intensity in a region as the constraint and thus are only applied to closed contours. The proposed snake model also takes into account the contour orientation so that any unrelated edges in the image will be discarded even if these edges have high gradient, or enclose a homogeneous region. Dynamic programming is used as the optimization method in our implementation and the im age information update is naturally incorporated in the op timization process. Experiment results on face edge and human tongue tracking are also presented in this paper and the robustness and accuracy of the proposed model is veri fied by quantitative and qualitative analysis.
Important Links:
Go Back
