Salient object detection based on Laplacian similarity metrics

被引:5
作者
Wang, Baoyan [1 ]
Zhang, Tie [2 ]
Wang, Xingang [3 ]
机构
[1] Northeastern Univ, Coll Informat Sci & Engn, 3-11 Wenhua Rd, Shenyang 110819, Liaoning, Peoples R China
[2] Northeastern Univ, Coll Sci, 3-11 Wenhua Rd, Shenyang 110819, Liaoning, Peoples R China
[3] Northeastern Univ Qinhuangdao, Sch Control Engn, 143 Taishan Rd, Qinhuangdao 066004, Peoples R China
基金
中国国家自然科学基金;
关键词
Superpixels; Saliency; Background prior; Similarity metrics; IMAGE; INTEGRATION; MODEL;
D O I
10.1007/s00371-017-1404-7
中图分类号
TP31 [计算机软件];
学科分类号
081202 ; 0835 ;
摘要
Background prior has become a novel viewpoint and made progresses in salient object detection. Most existing salient object detection algorithms based on background prior take boundaries as backgrounds and neglect nonbackground factors of boundaries, which is in fact unreasonable. Thus it is necessary to combine background prior with the analysis of boundary property. In this paper, the probability values computed by Mahalanobis distance are used to describe the likelihood of boundary superpixels belonging to backgrounds, which is viewed as a method for analyzing boundary properties. Meanwhile, some cues should be integrated with the obtained probability values for saliency computation. Inspired by the theory of Laplacian similarity metrics, two-stage complementary metrics are established according to different clusters in which two-stage queries lie, and a two-stage detection algorithm (SLSM) of salient objects is thus proposed by combining two-stage complementary similarity metrics with the probability values. Furthermore, when the detailed clusters (dense or sparse) of queries in each detection stage are ignored, an additional unified similarity metric is also constructed. Through the combination of the unified similarity metric and the proposed method for analyzing the boundary properties, another baseline algorithm (SLSMU) is also created. The results of experiments in which these two proposed algorithms are applied to four datasets demonstrate each of the two algorithms outperforms some existing state-of-the-art methods in terms of the different metrics.
引用
收藏
页码:645 / 658
页数:14
相关论文
共 44 条
  • [1] SLIC Superpixels Compared to State-of-the-Art Superpixel Methods
    Achanta, Radhakrishna
    Shaji, Appu
    Smith, Kevin
    Lucchi, Aurelien
    Fua, Pascal
    Suesstrunk, Sabine
    [J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2012, 34 (11) : 2274 - 2281
  • [2] Achanta R, 2009, PROC CVPR IEEE, P1597, DOI 10.1109/CVPRW.2009.5206596
  • [3] [Anonymous], 2007, PROC IEEE C COMPUT V, DOI 10.1109/CVPR.2007.383267
  • [4] Salient Object Detection: A Benchmark
    Borji, Ali
    Cheng, Ming-Ming
    Jiang, Huaizu
    Li, Jia
    [J]. IEEE TRANSACTIONS ON IMAGE PROCESSING, 2015, 24 (12) : 5706 - 5722
  • [5] Duan LJ, 2011, PROC CVPR IEEE, P473, DOI 10.1109/CVPR.2011.5995676
  • [6] Visual saliency estimation by nonlinearly integrating features using region covariances
    Erdem, Erkut
    Erdem, Aykut
    [J]. JOURNAL OF VISION, 2013, 13 (04):
  • [7] Context-Aware Saliency Detection
    Goferman, Stas
    Zelnik-Manor, Lihi
    Tal, Ayellet
    [J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2012, 34 (10) : 1915 - 1926
  • [8] A Novel Multiresolution Spatiotemporal Saliency Detection Model and Its Applications in Image and Video Compression
    Guo, Chenlei
    Zhang, Liming
    [J]. IEEE TRANSACTIONS ON IMAGE PROCESSING, 2010, 19 (01) : 185 - 198
  • [9] Image Signature: Highlighting Sparse Salient Regions
    Hou, Xiaodi
    Harel, Jonathan
    Koch, Christof
    [J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2012, 34 (01) : 194 - 201
  • [10] Automatic foveation for video compression using a neurobiological model of visual attention
    Itti, L
    [J]. IEEE TRANSACTIONS ON IMAGE PROCESSING, 2004, 13 (10) : 1304 - 1318