We propose a novel external force for active contours, which we call neighborhood-extending and noise-smoothing\r\ngradient vector flow (NNGVF). The proposed NNGVF snake expresses the gradient vector flow (GVF) as a\r\nconvolution with a neighborhood-extending Laplacian operator augmented by a noise-smoothing mask. We find\r\nthat the NNGVF snake provides better segmentation than the GVF snake in terms of noise resistance, weak edge\r\npreservation, and an enlarged capture range. The NNGVF snake accomplishes this with a reduced computational\r\ncost while maintaining other desirable properties of the GVF snake, such as initialization insensitivity and good\r\nconvergences at concavities. We demonstrate the advantages of NNGVF on synthetic and real images.
Loading....