A Multiscale Superpixel-Level Salient Object Detection Model Using Local-Global Contrast Cue

doi:10.1007/s12204-017-1810-z

Journal of shanghai Jiaotong University (Science) ›› 2017, Vol. 22 ›› Issue (1): 121-128.doi: 10.1007/s12204-017-1810-z

A Multiscale Superpixel-Level Salient Object Detection Model Using Local-Global Contrast Cue

MU Nan1 (穆楠), XU Xin1,2* (徐新), WANG Yinglin3 (王英林), ZHANG Xiaolong1,2 (张晓龙)

(1. School of Computer Science and Technology, Wuhan University of Science and Technology, Wuhan 430065, China; 2. Hubei Province Key Laboratory of Intelligent Information Processing and Real-Time Industrial System, Wuhan University of Science and Technology, Wuhan 430065, China; 3. School of Information Management and Engineering, Shanghai University of Finance and Economics, Shanghai 200433, China)
(1. School of Computer Science and Technology, Wuhan University of Science and Technology, Wuhan 430065, China; 2. Hubei Province Key Laboratory of Intelligent Information Processing and Real-Time Industrial System, Wuhan University of Science and Technology, Wuhan 430065, China; 3. School of Information Management and Engineering, Shanghai University of Finance and Economics, Shanghai 200433, China)

Online:2017-02-28 Published:2017-04-04
Contact: XU Xin1,2* (徐新) E-mail:xuxin0336@163.com

Abstract

Abstract: The goal of salient object detection is to estimate the regions which are most likely to attract human’s visual attention. As an important image preprocessing procedure to reduce the computational complexity, salient object detection is still a challenging problem in computer vision. In this paper, we proposed a salient object detection model by integrating local and global superpixel contrast at multiple scales. Three features are computed to estimate the saliency of superpixel. Two optimization measures are utilized to refine the resulting saliency map. Extensive experiments with the state-of-the-art saliency models on four public datasets demonstrate the effectiveness of the proposed model.

Key words: salient object detection| superpixel| multiple scales| local contrast| global contrast

摘要： The goal of salient object detection is to estimate the regions which are most likely to attract human’s visual attention. As an important image preprocessing procedure to reduce the computational complexity, salient object detection is still a challenging problem in computer vision. In this paper, we proposed a salient object detection model by integrating local and global superpixel contrast at multiple scales. Three features are computed to estimate the saliency of superpixel. Two optimization measures are utilized to refine the resulting saliency map. Extensive experiments with the state-of-the-art saliency models on four public datasets demonstrate the effectiveness of the proposed model.

关键词: salient object detection| superpixel| multiple scales| local contrast| global contrast

CLC Number:

TP 391.4

MU Nan1 (穆楠), XU Xin1,2* (徐新), WANG Yinglin3 (王英林), ZHANG Xiaolong1,2 (张晓龙). A Multiscale Superpixel-Level Salient Object Detection Model Using Local-Global Contrast Cue[J]. Journal of shanghai Jiaotong University (Science), 2017, 22(1): 121-128.

References 25

[1]	ITTI L, KOCH C, NIEBUR E. A model of saliencybasedvisual attention for rapid scene analysis [J].IEEE Transactions on Pattern Analysis and MachineIntelligence, 1998, 20(11): 1254-1259.
[2]	HAREL J, KOCH C, PERONA P. Graph-based visualsaliency [C]//Proceedings of Neural Information ProcessingSystems. Piscataway, NJ: IEEE, 2007: 545-552.
[3]	RIGAS I, ECONOMOU G, FOTOPOULOS S. Efficientmodeling of visual saliency based on local sparserepresentation and the use of hamming distance [J].Computer Vision and Image Understanding, 2015,134: 33-45.
[4]	ZHANG J, WANG M, ZHANG S, et al. Spatiochromaticcontext modeling for color saliency analysis [J].IEEE Transactions on Neural Networks and LearningSystems, 2016, 27(6): 1177-1189.
[5]	NOURI A, CHARRIER C, LEZORAY O. Multi-scalemesh saliency with local adaptive patches for viewpointselection [J]. Signal Processing: Image Communication,2015, 38: 151-166.
[6]	LUO W, LI H, LIU G, et al. Global salient informationmaximization for saliency detection [J]. Signal Processing:Image Communication, 2012, 27(3): 238-248.
[7]	WU J, SHEN X, ZHU W, et al. Mesh saliency withglobal rarity [J]. Graphical Models, 2013, 75(5): 255-264.
[8]	XU L, ZENG L, DUAN H. An effective vector modelfor global-contrast-based saliency detection [J]. Journalof Visual Communication and Image Representation,2015, 30: 64-74.
[9]	YEH H H, LIU K H, CHEN C S. Salient object detectionvia local saliency estimation and global homogeneityrefinement [J]. Pattern Recognition, 2014, 47(4):1740-1750.
[10]	TONG N, LU H, ZHANG Y, et al. Salient object detectionvia global and local cues [J]. Pattern Recognition,2015, 48(10): 3258-3267.
[11]	FAN Q, QI C. Saliency detection based on global andlocal short-term sparse representation [J]. Neurocomputing,2016, 175: 81-89.
[12]	ACHANTA R, SHAJI A, SMITH K, et al. SLIC superpixelscompared to state-of-the-art superpixel methods[J]. IEEE Transactions on Pattern Analysis andMachine Intelligence, 2012, 34(11): 2274-2282.
[13]	HE K, SUN J, TANG X. Single image haze removalusing dark channel prior [J]. IEEE Transactionson Pattern Analysis and Machine Intelligence, 2011,33(12): 2341-2353.
[14]	LIU T, YUAN Z, SUN J, et al. Learning to detecta salient object [J]. IEEE Transactions on PatternAnalysis and Machine Intelligence, 2011, 33(2): 353-367.
[15]	ALPERT S, GALUN M, BRANDT A, et al. Imagesegmentation by probabilistic bottom-up aggregationand cue integration [J]. IEEE Transactions on PatternAnalysis and Machine Intelligence, 2012, 34(2): 315-327.
[16]	YAN Q, XU L, SHI J, et al. Hierarchical saliency detection[C]//Proceedings of IEEE Conference on ComputerVision and Pattern Recognition. Piscataway, NJ:IEEE, 2013: 1155-1162.
[17]	YANG C, ZHANG L. LU H, et al. Saliency detectionvia graph-based manifold ranking [C]//Proceedings ofIEEE Conference on Computer Vision and PatternRecognition. Piscataway, NJ: IEEE, 2013: 3166-3137.
[18]	MURRAY N, VANRELL M, OTAZU X, et al. Saliencyestimation using a non-parametric low-level visionmodel [C]//Proceedings of IEEE Conference on ComputerVision and Pattern Recognition. Piscataway, NJ:IEEE, 2011: 433-440.
[19]	HOU X, HAREL J, KOCH C. Image signature: Highlightingsparse salient regions [J]. IEEE Transactionson Pattern Analysis and Machine Intelligence, 2012,34(1): 194-201.
[20]	GOFERMAN S, ZELNIK-MANOR L, TAL A.Context-aware saliency detection [J]. IEEE Transactionson Pattern Analysis and Machine Intelligence,2012, 34(10): 1915-1926.
[21]	SHEN X, WU Y. A unified approach to salientobject detection via low rank matrix recovery[C]//Proceedings of IEEE Conference on ComputerVision and Pattern Recognition. Piscataway, NJ:IEEE, 2012: 853-860.
[22]	MARGOLIN R, TAL A, ZELNIK-MANOR L. Whatmakes a patch distinct? [C]//Proceedings of IEEEConference on Computer Vision and Pattern Recognition.Piscataway, NJ: IEEE, 2013: 1139-1146.
[23]	ZHU W, LIANG S, WEI Y, et al. Saliencyoptimization from robust background detection[C]//Proceedings of IEEE Conference on ComputerVision and Pattern Recognition. Piscataway, NJ:IEEE, 2014: 2814-2821.
[24]	TONG N, LU H, ZHANG L, et al. Saliency detectionwith multi-scale superpixels [J]. IEEE Signal ProcessingLetters, 2014, 21(9): 1035-1039.
[25]	TONG N, LU H, RUAN X, et al. Salient object detectionvia bootstrap learning [C]//Proceedings of IEEEConference on Computer Vision and Pattern Recognition.Piscataway, NJ: IEEE, 2015: 1884-1892.

A Multiscale Superpixel-Level Salient Object Detection Model Using Local-Global Contrast Cue

A Multiscale Superpixel-Level Salient Object Detection Model Using Local-Global Contrast Cue

Knowledge

Abstract

Cite this article

share this article

References 25

Related Articles 1

Recommended Articles

Metrics

Comments