A Modified Selective Attention Model for Salient Region Detection in Real Scenes
One of the most important and essential parts of image processing tasks in computer vision is to predict the regions of interest that is the most attractive and also representative salient ones. This task can be realized effortless by the human visual system via the function of selective attention. In this paper, we introduced a set of new visual features, including contract, entropy, and local feature change into the Itti bottom-up model. We used a set of qualitative and quantitative analysis to demonstrate the effectiveness of the proposed approach on a large dataset with different scenes, and the results show that the modified Itti model combined with the new features can improve greatly the efficiency of salient region detection in real scenes.
Keywordsvisual attention bottom-up attention model feature extraction contract entropy intensity change
Unable to display preview. Download preview PDF.
- 1.Koch, C., Ullman, S.: Shifts in selective visual attention: towards the underlying neural circuitry. Hum. Neurobiol. 4(4), 219–227 (1985)Google Scholar
- 3.Walther, D., Koch, C.: Saliency Toolbox 2.0. (2006)Google Scholar
- 4.Hou, X., Zhang, L.: Saliency detection: A spectral residual approach. In: CVPR, pp. 1–8. IEEE (2007)Google Scholar
- 5.Bruce, N., Tsotsos, J.: Saliency based on information maximization. In: Advances in Neural Information Processing Systems, vol. 18, p. 155 (2006)Google Scholar
- 6.Harel, J., Koch, C., Perona, P.: Graph-based visual saliency. In: Advances in Neural Information Processing Systems, vol. 19, p. 545 (2007)Google Scholar
- 12.Judd, T., Ehinger, K., Durand, F., Torralba, A.: Learning to predict where humans look. In: ICCV, pp. 2106–2113. IEEE (2009)Google Scholar
- 13.Kootstra, G., Nederveen, A., De Boer, B.: Paying attention to symmetry. In: Proceedings of the British Machine Vision Conference, pp. 1115–1125 (2008)Google Scholar