Salient object detection by robust foreground and background seed selection. (March 2021)

• Record Type: Journal Article
• Authors: Wang, Huibin; Zhu, Chao; Shen, Jie; Zhang, Zhen; Shi, Xiaotao
• Is Part Of: Computers & electrical engineering. Volume 90(2021)
• Publication Date: 2021-03
• Subjects: Salient object detection -- Background seed selection -- Foreground seed selection -- Boundary prior -- Image feature -- Convex hull; Computer engineering -- Periodicals; Electrical engineering -- Periodicals; Electrical engineering -- Data processing -- Periodicals; Ordinateurs -- Conception et construction -- Périodiques; Électrotechnique -- Périodiques; Électrotechnique -- Informatique -- Périodiques; Computer engineering; Electrical engineering; Electrical engineering -- Data processing; Periodicals; Electronic journals; 621.302854
• DOI: 10.1016/j.compeleceng.2021.106993
• Languages: English
• ISSNs: 0045-7906

Abstract

Highlights: A novel seed selection method based on multiple prior knowledge and joint image features is proposed to contribute to salient object detection. Break through the limitations of the boundary prior, and accurately select background seed. Effectively locate the salient object, and combine multiple image features to quickly select foreground seed. Fully explore the saliency information of image foreground and background. Abstract: This paper presents a novel background and foreground seed selection method for graph-based salient object detection. First, according to the boundary prior which considers that the image boundary is mainly the background, we select the initial background seed set and optimize it through our proposed two-stage background seed correction processes by combing multiple internal image features. Second, different from most existing center-based prior methods, this paper uses the convex hull of the point of interest to estimate the location of salient objects and thus obtains the initial foreground seed set. Moreover, foreground seeds are refined by the proposed foreground seed correction process, which depends on the color and spatial differences between seeds. Third, we adopt the extended random walk to propagate the background and foreground labels. Finally, a fusion model is proposed to integrate the background- and foreground-based saliency maps, generating final salient object detection results. Experiments on publicly available data sets show that the proposed algorithm achieves better results in contrast to other state-of-the-art methods. Graphical abstract: Image, graphical abstract