《计算机应用研究》|Application Research of Computers

基于马尔可夫场的图像背景拼贴

Photo background stitching based on Markov random field

免费全文下载 (已被下载 次)  
获取PDF全文
作者 王诗尧,胡涛
机构 武汉大学 计算机学院,武汉 430072
统计 摘要被查看 次,已被下载
文章编号 1001-3695(2017)09-2849-05
DOI 10.3969/j.issn.1001-3695.2017.09.063
摘要 针对现代消费摄像机视野较小导致难以拍摄全景图片的问题,提出了一种新的图像背景拼贴的方法。给定一幅被扩展的输入图像,从网络上获取包含相关场景的散乱图片集,重建出三维模型后对每幅图片颜色迁移,使得每张图片与输入图像的颜色基调保持一致;在重建的三维模型的基础上,将超像素分割后的候选集图片都形变到输入图像的视角;建立一个马尔可夫随机场模型,使用带标签的图割方法求解模型完成图像融合。实验结果表明,该算法适用于拥有大量网络图片集的大多数场景。与现有算法相比,该算法能得到与真实场景更为相符的结果,并且拼接效果更为自然。
关键词 超像素;三维重建;马尔可夫随机场;图割
基金项目 国家自然科学基金资助项目(61562025)
本文URL http://www.arocmag.com/article/01-2017-09-063.html
英文标题 Photo background stitching based on Markov random field
作者英文名 Wang Shiyao, Hu Tao
机构英文名 SchoolofComputer,WuhanUniversity,Wuhan430072,China
英文摘要 Concerning the limited field view of most consumer cameras that leads to the difficulty of capturing large scene, this paper proposed a new method to solve this problem, which was able to stitch image’s background effectively. Input was a li-mited view image. First, it gathered the scattered photos from the Internet with each photograph changed into same color tone using color convert method. Then, these collected photos were cut into super-pixels and then warped to the same view of the input image on the reconstructed 3D model basis. Finally, it built a novel Markov random field based model to composite images. It solved the energy function with Markov random field by graph cuts with label costs. The experimental results of this method is more close to the real-world landmarks, and provides convincing visual context which is real spatial and not available in the user images than existing works.
英文关键词 super-pixel; 3D reconstruction; Markov random field; graph cut
参考文献 查看稿件参考文献
  [1] Garg R, Seitz M. Dynamic mosaics[C] //Proc of the 2nd Internatio-nal Conference on 3D Imaging, Modeling, Processing, Visualization and Transmission. Washington DC:IEEE Computer Society, 2012:65-72.
[2] Zelnik-Manor L, Peters G, Perona P. Squaring the circle in panoramas[C] //Proc of IEEE International Conference on Computer Vision. 2005:1292-1299.
[3] Richardt C, Pritch Y, Zimmer H, et al. Megastereo:constructing high resolution stereo panoramas[C] //Proc of IEEE Conference on Compu-ter Vision and Pattern Recognition. Washington DC:IEEE Computer Society, 2013:1256-1263.
[4] Wang Miao, Lai Yukun, Liang Yuan, et al. BiggerPicture:data-driven image extrapolation using graph matching[J] . ACM Trans on Graphics, 2014, 33(6):1-13.
[5] Whyte O, Sivic J, Zisserman A. Get out of my picture! Internet-based inpainting[C] // Proc of the 20th British Machine Vision Conference. 2009:125-132.
[6] Zhang Runze, Li Shiwei, Fang Tian, et al. Joint camera clustering and surface segmentation for large-scale multi-view stereo[C] //Proc of IEEE International Conference on Computer Vision. Washington DC:IEEE Computer Society, 2015:2084-2092.
[7] Bódis-Szomorú A, Riemenschneider H, Van Gool L. Superpixel meshes for fast edge-preserving surface reconstruction[C] //Proc of IEEE Conference on Computer Vision and Pattern Recognition. 2015:2011-2020.
[8] Zhang Chenxi, Gao Jizhou, Wang O, et al. Personal photo enhancement using Internet photo collections[J] . IEEE Trans on Visualization and Computer Graphic, 2014, 20(2):262-275.
[9] Wu Changchang. VisualSFM :a visual structure from motion system[EB/OL] . http://ccwu. me/vsfm/.
[10] Reinhard E, Ashikhmin M, Gooch B, et a1. Color transfer between images[J] . IEEE Computer Graphics and Applications, 2001, 21(5):34-41.
[11] Achanta R, Shaji A, Smith K. SLIC superpixels compared to state-of-the-art superpixwl methods[J] . IEEE Trans on Pattern Analysis and Machine Intelligence, 2012, 34(11):2274-2282.
[12] Efros A A, Freeman W T. Image quilting for texture synthesis and transfer[C] //Proc of the 28th Annual Conference on Computer Graphics and Internative Techniques. 2001:341-346.
[13] Nomura Y, Zhang Li, Nayar S K. Scene collages and flexible camera arrays[C] // Proc of the 18th Eurographics Conference on Rendering Techniques. 2007:127-138.
[14] Liu Feng, Gleicher M, Jin Hailin, et al. Content-preserving warps for 3D video stabilization[J] . IEEE Trans on Pattern Analysis and Machine Intelligence, 2009, 28(3):1-9.
[15] Chaurasia G, Sorkine O, Drettakis G. Silhouette-aware warping for image-based rendering[J] . Computer Graphics Forum, 2011, 30(4):1223-1232.
[16] Delong A, Osokin A, Isack H, et al. Fast approximate energy minimization via graph cuts[J] . IEEE Trans on Pattern Analysis and Machine Intelligence, 20(12):1222-1239.
[17] Kolomogoro V, Zabih R. What energy functions can be minimized via graph cuts[J] . IEEE Trans on Pattern Analysis and Machine Intelligence, 2004, 26(2):147-159.
[18] Boykov Y, Kolmogorov V. An experimental comparison of min-cut/max-flow algorithms for energy minimization in vision[J] . IEEE Trans on Pattern Analysis and Machine Intelligence, 2004, 26(9):1124-1137.
[19] Delong A, Osokin A, Isack H, et al. Fast approximate energy minimization with label costs[C] //Proc of IEEE Conference on Computer Vision and Pattern Recognition. Berlin:Springer, 2010:1-27.
收稿日期 2016/7/1
修回日期 2016/8/11
页码 2849-2853
中图分类号 TP391.41
文献标志码 A