Saliency-guided level set model for automatic object segmentation. (September 2019)

• Record Type: Journal Article
• Authors: Cai, Qing; Liu, Huiying; Qian, Yiming; Zhou, Sanping; Duan, Xiaojun; Yang, Yee-Hong
• Is Part Of: Pattern recognition. Volume 93(2019:Sep.)
• Page Start: 147
• Page End: 163
• Publication Date: 2019-09
• Subjects: Level set model -- Object segmentation -- Visual saliency -- Graph cuts -- Automatic initialization; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2019.04.019
• Languages: English
• ISSNs: 0031-3203

Abstract

Highlights: A global saliency-guided energy term is defined using the saliency map to roughly extract the objects, which significantly improves the segmentation efficiency and the robustness to noise and to initialize the SLSM. Unlike most existing level set models only using grayscale information of color images, the proposed local multichannel-based energy term using the CIEL*a*b* color space successfully achieves color image segmentation. A novel graph cuts based method is proposed using the Heaviside function of the level set model to define the data term that can avoid the occurrence of small isolated region in the final segmentation. A new automatic initialization method using graph cuts to segment the image saliency map avoids the tedious and time-consuming manual initialization and further improves the segmentation efficiency. Abstract: The level set model is a popular method for object segmentation. However, most existing level set models perform poorly in color images since they only use grayscale intensity information to defined their energy functions. To address this shortcoming, in this paper, we propose a new saliency-guided level set model (SLSM), which can automatically segment objects in color images guided by visual saliency. Specifically, we first define a global saliency-guided energy term to extract the color objects approximately. Then, by integrating information from different color channels, we define a novel local multichannel based energy term to extract the color objects in detail. In addition, unlike using a length regularization term in the conventional level set models, we achieve segmentation smoothness by incorporating our SLSM into a graph cuts formulation. More importantly, the proposed SLSM is automatically initialized by saliency detection. Finally, the evaluation on public benchmark databases and our collected database demonstrates that the new SLSM consistently outperforms many state-of-the-art level set models and saliency detecting methods in accuracy and robustness.