A survey of thresholding techniques
Computer Vision, Graphics, and Image Processing
Region Competition: Unifying Snakes, Region Growing, and Bayes/MDL for Multiband Image Segmentation
IEEE Transactions on Pattern Analysis and Machine Intelligence
International Journal of Computer Vision
A Multiphase Level Set Framework for Image Segmentation Using the Mumford and Shah Model
International Journal of Computer Vision
Using Prior Shapes in Geometric Active Contours in a Variational Framework
International Journal of Computer Vision
Level Set Evolution without Re-Initialization: A New Variational Formulation
CVPR '05 Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) - Volume 1 - Volume 01
Color Image Segmentation Based on Mean Shift and Normalized Cuts
IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics
A topology preserving level set method for geometric deformable models
IEEE Transactions on Pattern Analysis and Machine Intelligence
IEEE Transactions on Image Processing
Hi-index | 0.00 |
In most existing level set models for image segmentation, it is necessary to constantly re-initialize the level set function, or to acquire the gradient flow information of the image to restrict the evolution of the curve. A novel image segmentation model of level set is proposed in the paper, which is based on the maximization of the between-class variance and the distance-based constraint function. In this model, the distance-based constraint function is introduced as the internal energy to ensure that the level set function is always the signed distance function (SDF), so that the constant re-initialization of the level set function during the evolution process is avoided. Meanwhile, the external energy function (between-class variance function) is constructed based on the weighted sum of square of the difference between the average grey levels of the target region and the overall region, the background and the overall region respectively. This function is maximized to ensure that the curve represented by zero level set converges towards the target boundary stably. Experimental results show that the constant re-initialization in traditional models has been eliminated in the proposed model. Furthermore, since region information has been incorporated into the energy function, the model renders good performance in the segmentation of both weak edges images and those with Gaussian noise or impulse noise.