Multiphase Tensor Level-Set Method for Segmentation of Natural Images
Methods for processing digital images based on partial differential equations (PDE) have been intensively developed in image analysis since 1990s. Recently presented works which follow this approach show good results in image segmentation. In this work we present a novel method based on a tensor level-set approach. By using superpixel oversegmentation as one of the image cues we simplify evolution of level-set equation. This approach allows us to extend tensor level-set to multiphase tensor-level set method. To make our model applicable on color images we use cues obtained from CIELAB color space. Evaluation results show that our method achieves better segmentation results than the existing models based on level-set framework.
Unable to display preview. Download preview PDF.
- 3.Osher, S.J., Fedkiw, R.: Level Set Methods and Dynamic Implicit Surfaces. Springer, Heidelberg (2002)Google Scholar
- 7.Gibou, F., Fedkiw, R.: A fast hybrid k-means level set algorithm for segmentation. In: 4th Annual Hawaii International Conference on Statistics and Mathematics, pp. 281–291. Citeseer (2005)Google Scholar
- 8.Bertelli, L., Sumengen, B., Manjunath, B.S., Gibou, F.: A variational framework for multiregion pairwise-similarity-based image segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1400–1414 (2007)Google Scholar
- 9.Martin, D.R., Fowlkes, C.C., Tal, D., Malik, J.: A database of human segmented natural images and its application to evaluating segmentation algorithms. In: Workshop on Perceptual Organization in Computer Vision (2001)Google Scholar
- 10.Ren, X.F., Malik, J.: Learning a classification model for segmentation. In: ICCV, pp. 10–17 (2003)Google Scholar
- 11.Fowlkes, C., Arbeláez, P., Maire, M., Malik, J.: Contour Detection and Hierarchical Image Segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence (2010)Google Scholar