Shadow Detection Based on Graph Cuts for a Single Image
-
摘要: 为了准确检测单幅图像中的影子, 提出一种基于图割的影子检测方法. 首先,使用均值漂移将原始图像分割为若干区域并记录区域之间的边界. 其次,利用支持向量机分类器分别获得分割图像中的候选影子边界和候选影子非影子区域对. 然后,利用候选影子边界两侧的区域信息及候选影子非影子区域对信息构造一个能量函数, 该能量函数反映了将图像中一部分区域划分为影子区域而另一部分区域划分为非影子区域时所需的代价. 再次,结合该能量函数构造出无向图,并证明所构造的无向图的最小割对应能量函数的最小值. 最后,通过图割算法求解该能量函数得到最终的影子检测结果. 实验结果表明,与现有代表最新进展的单幅图像影子检测方法相比,所提方法提高了影子检测结果的准确性和连续性.Abstract: In order to detect the shadow in a single image accurately, a shadow detection approach based on graph cuts for a single image is proposed in this paper. Firstly, the original image is segmented into several regions using the Mean-Shift algorithm, and the boundary information between adjacent regions is recorded. Secondly, the candidate shadow boundary and the candidate shadow-nonshadow region pair are obtained respectively by using the support vector machine classifier. Then, an energy function, which reflects the cost of dividing some image regions as shadow regions and the others as nonshadow ones, is constructed by utilizing regions' information on both sides of the candidate shadow boundary and the candidate shadow-nonshadow region pair. Furthermore, combining with the energy function, an undirected graph is constructed and it is proved that the minimum cut of the graph corresponds to the minimum of the energy function. Finally, the energy function is solved with the graph cuts algorithm and the final shadow regions in an image are gained. The experimental results show that, compared with the latest shadow detection methods for a single image, the proposed approach improves the accuracy and continuity of the results.
-
Key words:
- Outdoor image /
- shadow detection /
- graph cuts /
- boundary information /
- region information
-
[1] Al-Najdawi N, Bez H E, Singhai J, Edirisinghe E A. A survey of cast shadow detection algorithms. Pattern Recognition Letters, 2012, 33(6): 752-764 [2] [2] Adeline K R M, Chen M, Briottet X, Pang S K, Paparoditis N. Shadow detection in very high spatial resolution aerial images: a comparative study. ISPRS Journal of Photogrammetry and Remote Sensing, 2013, 80(6): 21-38 [3] [3] Sunkavalli K, Zickler T, Pfister H. Visibility subspaces: uncalibrated photometric stereo with shadows. In: Proceedings of the 11th European Conference on Computer Vision. Berlin, Heidelberg: Springer, 2011. 251-264 [4] [4] Karsch K, Hedau V, Forsyth D, Hoiem D. Rendering synthetic objects into legacy photographs. ACM Transactions on Graphics, 2011, 30(6): 1-12 [5] [5] Abrams A, Miskell K, Pless R. The Episolar constraint: monocular shape from shadow correspondence. In: Proceedings of the 2013 IEEE Conference on Computer Vision and Pattern Recognition. Portland, OR: IEEE, 2013. 1407-1414 [6] [6] Finlayson G D, Hordley S D, Lu C, Drew M S. On the removal of shadows from images. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2006, 28(1): 59-68 [7] [7] Maxwell B A, Friedhoff R M, Smith C A. A bi-illuminant dichromatic reflection model for understanding images. In: Proceedings of the 26th IEEE Conference on Computer Vision and Pattern Recognition. Anchorage, AK: IEEE, 2008. 1-8 [8] [8] Tian J D, Sun J, Tang Y D. Tricolor attenuation model for shadow detection. IEEE Transactions on Image Processing, 2009, 18(10): 2355-2363 [9] [9] Zhu J J, Samuel K G G, Mashood S Z, Tappen M F. Learning to recognize shadows in monochromatic natural images. In: Proceedings of the 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. San Francisco, CA: IEEE, 2010. 223-230 [10] Lalonde J F, Efros A A, Narasimhan S G. Detecting ground shadows in outdoor consumer photographs. In: Proceedings of the 11th European Conference on Computer Vision. Berlin, Heidelberg: Springer, 2010. 322-335 [11] Huang X, Hua G, Tumblin J, Williams L. What characterizes a shadow boundary under the sun and sky. In: Proceedings of the 2011 IEEE International Conference on Computer Vision. Barcelona, Spain: IEEE, 2011. 898-905 [12] Guo R Q, Dai Q Y, Hoiem D. Paired regions for shadow detection and removal. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013, 35(12): 2956-2967 [13] Martin D R, Fowlkes C C, Malik J. Learning to detect natural image boundaries using local brightness, color, and texture cues. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2004, 26(5): 530-549 [14] Dalal N, Triggs B. Histograms of oriented gradients for human detection. In: Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. San Diego, CA: IEEE, 2005. 886-893 [15] Boykov Y, Veksler O, Zabih R. Fast approximate energy minimization via graph cuts. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2001, 23(11): 1222- 1239
点击查看大图
计量
- 文章访问数: 1359
- HTML全文浏览量: 64
- PDF下载量: 668
- 被引次数: 0