|Table of Contents|

Image saliency and intensity inhomogeneity based automatic object extraction


Research Field:
Publishing date:


Image saliency and intensity inhomogeneity based automatic object extraction
Guo DongyanZhao ChunxiaLi JunxiaDing Jundi
School of Computer Science and Engineering,NUST,Nanjing 210094,China
image saliency intensity inhomogeneity image segmentation object extraction
A novel automatic object extraction method is proposed based on the pixel intensity inhomogeneity and the image saliency.Firstly,two kinds of seeds with different inhomogeneities are determined by exploring the pixel inhomogeneity factor(PIF)and the neighborhood inhomogeneity factor(NIF).Secondly,several equivalence classes are formed by seeds growing based on equivalence partitioning.Over-segmentation result of the original image is then obtained after adding the noise points to the nearest equivalence classes.At last,the interested object is extracted by combining the over-segmentation result with the image saliency detection technique.In the method,the intensity inhomogeneity information of pixels is considered and used in image segmentation for the first time.Experimental results demonstrate the effectiveness and robustness of the proposed method in automatically extracting interested objects.


[1] Chen L,Xie X,Fan X,et al.A visual attention model for adapting images on small displays[J].Multimedia Systems,2003,9(4):353-364.
[2]Nicolas P,David D,James J.Why is real-world visual object recognition hard?[J].PLoS Computational Biology,2008,4(1):151-156.
[3]Mignotte M.Segmentation by fusion of histogram-based k-means clusters in different color spaces[J].IEEE Transactions on Image Processing,2008,17(5):780-787.
[4]Cai W,Chen S,Zhang D.Fast and robust fuzzy c-means clustering algorithms incorporating local information for image segmentation[J].Pattern Recognition,2007,40(3):825-838.
[5]Comaniciu D,Meer P.Mean shift:a robust approach toward feature space analysis[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2002,24(5):603-619.
[6]Cousty J,Bertrand G,Najman L,et al.Watershed cuts:minimum spanning forests and the drop of water principle[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2009,31(8):1362-1374.
[7]Shi J,Malik J.Normalized cuts and image segmentation[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22(8):888-905.
[8]Ding J,Shen J,Pang H,et al.Exploiting intensity inhomogeneity to extract textured objects from natural scenes[A].Asian Conference on Computer Vision(Part Ⅲ),Lecture Notes in Computes Science[C].Berlin,Germany:Springer,2009,5996:1-10.
[9]周赟,李久贤,夏良正.基于区域生长的红外图像分割[J].南京理工大学学报,2002,26(1):75-78. Zhou Yun,Li Jiuxian,Xia Liangzheng.Infrared image segmentation based on region growing[J].Journal of Nanjing University of Science and Technology,2002,26(1):75-78.
[10]Itti L,Kouch C,Niebur E.A model of saliency-based visual attention for rapid scene analysis[J].IEEE Transactions on Pattern Analysisand Machine Intelligence,1998,20(11):1254-1259.
[11]Achanta R,Hemami S,Estrada F,et al.Frequency-tunedsalient region detection[A].IEEE International Conference on Computer Vision and Pattern Recognition[C].Miami,US:IEEE,2009:1597-1604.
[12]Harel J,Koch C,Perona P.Graph-based visual saliency[A].Advancesin Neural Information Processing Systems[C].Vancouver,Canada:MIT Press,2007,19:545-552.
[13]Borji A,Itti L.Exploiting local and global patch rarities for saliency detection[A].IEEE International Conference on Computer Vision and Pattern Recognition[C].RI,US:IEEE,2012:478-485.
[14]辛月兰.基于Grabcut的图像目标提取[J].青海师范大学学报(自然科学版),2012,28(3):30-34. Xin Yuelan.Image target detection based on the Grabcut[J].Journal of Qinghai Normal University(Natural Science Edition),2012,28(3):30-34.
[15]Alpert S,Basri S,Brandt M.Image segmentationby probabilistic bottom-up aggregation and cue integration[A].IEEE International Conference on CVPR[C].Minnoapolis,US:IEEE,2007:1-8.


Last Update: 2014-10-31