单幅室外图像的高阶能量方程阴影检测算法
|
摘要
单幅室外图像的阴影检测是数字图像处理领域的研究热点之一.针对单幅室外图像的阴影检测算法往往只关注阴影的边缘或区域信息,忽视了两者之间的依存关系,即"阴影区域-阴影边缘-非阴影区域"的阴影布局信息的问题,提出阴影检测算法.首先将图像分割成独立区域;然后利用支持向量机构建高阶能量方程,对阴影布局信息进行建模;最后通过最小化方程来判断区域是否为阴影.文中不但严谨证明了高阶能量方程在局部最优点下的性质,而且指出了其最优的降阶方法.在公开图像数据库上的实验表明,该算法能够有效地检测单幅室外图像中的阴影区域.
Detecting shadows from a single outdoor image has been one of research hotspots in image processing. To solve the problem that existing algorithms often focused on either shadow edges or shadow areas and ignored their dependence, specifically a layout information of "shadow area-shadow edge-non-shadow area", this paper proposed a shadow detection algorithm: firstly, an outdoor image was segmented into regions; secondly, the layout information was modeled by constructing a high order energy function with the help of support vector machine; finally, shadows were determined by minimizing this function. This paper also rigorously proved the property of the proposed high order energy function under the local optimum, and introduced an optimal method of reducing itself into a second order energy function. Experiments on public image databases show that the proposed algorithm can effectively detect shadows from a single outdoor image.
Detecting shadows from a single outdoor image has been one of research hotspots in image processing. To solve the problem that existing algorithms often focused on either shadow edges or shadow areas and ignored their dependence, specifically a layout information of "shadow area-shadow edge-non-shadow area", this paper proposed a shadow detection algorithm: firstly, an outdoor image was segmented into regions; secondly, the layout information was modeled by constructing a high order energy function with the help of support vector machine; finally, shadows were determined by minimizing this function. This paper also rigorously proved the property of the proposed high order energy function under the local optimum, and introduced an optimal method of reducing itself into a second order energy function. Experiments on public image databases show that the proposed algorithm can effectively detect shadows from a single outdoor image.
引文
[1]Forsyth D A,Ponce J.Computer vision:a modern approach[M].Upper Saddle River:Prentice Hall,2003
[2]Sonka M,Hlavac,V,Boyle R.Image processing,analysis,and machine vision[M].Heidelberg:Springer,1999
[3]Finlayson G D,Hordley S D,Cheng L,et al.On the removal of shadows from images[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2006,28(1):59-68
[4]Finlayson G D,Drew M S,Cheng L.Intrinsic images by entropy minimization[C]//Proceedings of the European Conference on Computer Vision.Heidelberg:Springer,2004:582-595
[5]Finlayson G D,Hordley S D,Drew M S.Removing shadows from images[C]//Proceedings of the European Conference on Computer Vision.Heidelberg:Springer,2002:823-836
[6]Guo R Q,Dai Q Y,Hoiem D.Single-image shadow detection and removal using paired regions[C]//Proceedings of the IEEEConference on Computer Vision and Pattern Recognition.Los Alamitos:IEEE Computer Society Press,2011:2033-2040
[7]Joshi A J,Papanikolopoulos N P.Learning to detect moving shadows in dynamic environments[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2008,30(11):2055-2063
[8]Vicente T F Y,Yu C P,Samaras D.Single image shadow detection using multiple cues in a supermodular MRF[C]//Proceedings of the British Machine Vision Conference.Swansea:BMVC Press,2013:126.1-126.12
[9]Vicente T F Y,Hoai M,Samaras D.Leave-one-out kernel optimization for shadow detection and removal[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2018,40(3):682-695
[10]Shen L,Chua T W,Leman K.Shadow optimization from structured deep edge detection[C]//Proceedings of the IEEEConference on Computer Vision and Pattern Recognition.Los Alamitos:IEEE Computer Society Press,2015:2067-2074
[11]Khan S H,Bennamoun M,Sohel F,et al.Automatic shadow detection and removal from a single image[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2016,38(3):431-446
[12]Kohli P,Ladicky L,Torr P H S.Robust higher order potentials for enforcing label consistency[C]//Proceedings of the IEEEConference on Computer Vision and Pattern Recognition.Los Alamitos:IEEE Computer Society Press,2008:1-8
[13]Boros E,Hammer P L.Pseudo-boolean optimization[J].Discrete Applied Mathematics,2002,123(1-3):155-225
[14]Gr?tschel M,Lovász L,Schrijver A.The ellipsoid method and its consequences in combinatorial optimization[J].Combinatorica,1981,1(2):169-197
[15]Rother C,Kolmogorov V,Lempitsky V,et al.Optimizing binary MRFs via extended roof duality[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Los Alamitos:IEEE Computer Society Press,2007:1-8
[16]Kolmogorov V.Convergent tree-reweighted message passing for energy minimization[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2006,28(10):1568-1583
[17]Felzenszwalb P F,Huttenlocher D P.Efficient belief propagation for early vision[J].International Journal of Computer Vision,2006,70(1):41-54
[18]Gorelick L,Boykov Y,Veksler O,et al.Submodularization for binary pairwise energies[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Los Alamitos:IEEE Computer Society Press,2014:1154-1161
[19]Kolmogorov V,Zabih R.What energy functions can be minimized via graph cuts[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2004,26(2):147-159
[20]Freedman D,Drineas P.Energy minimization via graph cuts:settling what is possible[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Los Alamitos:IEEE Computer Society Press,2005:939-946
[21]Ishikawa H.Transformation of general binary MRF minimization to the first-order case[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2011,33(6):1234-1249
[22]Fix A,Gruber A,Boros E,et al.A graph cut algorithm for higher-order Markov Random Fields[C]//Proceedings of the IEEE International Conference on Computer Vision.Los Alamitos:IEEE Computer Society Press.2011:1020-1027
[23]Gallagher A C,Batra D,Parikh D.Inference for order reduction in Markov random fields[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Los Alamitos:IEEE Computer Society Press,2011:1857-1864
[24]Zhu J J,Samuel K G G,Masood S Z,et al.Learning to recognize shadows in monochromatic natural images[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Los Alamitos:IEEE Computer Society Press,2010:223-230
[25]Arbelaez P,Maire M,Fowlkes C,et al.Contour detection and hierarchical image segmentation[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2011,33(5):898-916
[26]Tian J D,Qi X J,Qu L Q,et al.New spectrum ratio properties and features for shadow detection[J].Pattern Recognition,2016,51:85-96
[27]Qu L Q,Tian J D,Fan H J,et al.Evaluation of shadow features[J].IET Computer Vision,2018,12(1):95-103
[2]Sonka M,Hlavac,V,Boyle R.Image processing,analysis,and machine vision[M].Heidelberg:Springer,1999
[3]Finlayson G D,Hordley S D,Cheng L,et al.On the removal of shadows from images[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2006,28(1):59-68
[4]Finlayson G D,Drew M S,Cheng L.Intrinsic images by entropy minimization[C]//Proceedings of the European Conference on Computer Vision.Heidelberg:Springer,2004:582-595
[5]Finlayson G D,Hordley S D,Drew M S.Removing shadows from images[C]//Proceedings of the European Conference on Computer Vision.Heidelberg:Springer,2002:823-836
[6]Guo R Q,Dai Q Y,Hoiem D.Single-image shadow detection and removal using paired regions[C]//Proceedings of the IEEEConference on Computer Vision and Pattern Recognition.Los Alamitos:IEEE Computer Society Press,2011:2033-2040
[7]Joshi A J,Papanikolopoulos N P.Learning to detect moving shadows in dynamic environments[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2008,30(11):2055-2063
[8]Vicente T F Y,Yu C P,Samaras D.Single image shadow detection using multiple cues in a supermodular MRF[C]//Proceedings of the British Machine Vision Conference.Swansea:BMVC Press,2013:126.1-126.12
[9]Vicente T F Y,Hoai M,Samaras D.Leave-one-out kernel optimization for shadow detection and removal[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2018,40(3):682-695
[10]Shen L,Chua T W,Leman K.Shadow optimization from structured deep edge detection[C]//Proceedings of the IEEEConference on Computer Vision and Pattern Recognition.Los Alamitos:IEEE Computer Society Press,2015:2067-2074
[11]Khan S H,Bennamoun M,Sohel F,et al.Automatic shadow detection and removal from a single image[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2016,38(3):431-446
[12]Kohli P,Ladicky L,Torr P H S.Robust higher order potentials for enforcing label consistency[C]//Proceedings of the IEEEConference on Computer Vision and Pattern Recognition.Los Alamitos:IEEE Computer Society Press,2008:1-8
[13]Boros E,Hammer P L.Pseudo-boolean optimization[J].Discrete Applied Mathematics,2002,123(1-3):155-225
[14]Gr?tschel M,Lovász L,Schrijver A.The ellipsoid method and its consequences in combinatorial optimization[J].Combinatorica,1981,1(2):169-197
[15]Rother C,Kolmogorov V,Lempitsky V,et al.Optimizing binary MRFs via extended roof duality[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Los Alamitos:IEEE Computer Society Press,2007:1-8
[16]Kolmogorov V.Convergent tree-reweighted message passing for energy minimization[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2006,28(10):1568-1583
[17]Felzenszwalb P F,Huttenlocher D P.Efficient belief propagation for early vision[J].International Journal of Computer Vision,2006,70(1):41-54
[18]Gorelick L,Boykov Y,Veksler O,et al.Submodularization for binary pairwise energies[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Los Alamitos:IEEE Computer Society Press,2014:1154-1161
[19]Kolmogorov V,Zabih R.What energy functions can be minimized via graph cuts[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2004,26(2):147-159
[20]Freedman D,Drineas P.Energy minimization via graph cuts:settling what is possible[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Los Alamitos:IEEE Computer Society Press,2005:939-946
[21]Ishikawa H.Transformation of general binary MRF minimization to the first-order case[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2011,33(6):1234-1249
[22]Fix A,Gruber A,Boros E,et al.A graph cut algorithm for higher-order Markov Random Fields[C]//Proceedings of the IEEE International Conference on Computer Vision.Los Alamitos:IEEE Computer Society Press.2011:1020-1027
[23]Gallagher A C,Batra D,Parikh D.Inference for order reduction in Markov random fields[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Los Alamitos:IEEE Computer Society Press,2011:1857-1864
[24]Zhu J J,Samuel K G G,Masood S Z,et al.Learning to recognize shadows in monochromatic natural images[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Los Alamitos:IEEE Computer Society Press,2010:223-230
[25]Arbelaez P,Maire M,Fowlkes C,et al.Contour detection and hierarchical image segmentation[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2011,33(5):898-916
[26]Tian J D,Qi X J,Qu L Q,et al.New spectrum ratio properties and features for shadow detection[J].Pattern Recognition,2016,51:85-96
[27]Qu L Q,Tian J D,Fan H J,et al.Evaluation of shadow features[J].IET Computer Vision,2018,12(1):95-103