摘要
针对当前内容感知的重定向方法中可能出现的变形和失真问题,提出一种融合显著与深度信息的缝切割重定向方法.首先利用GBVS算法获取图像显著信息,结合图像梯度信息与通过SIFT匹配方法获取的图像深度信息构建更精确的重要度图;其次,根据重要度图的能量分布,对原始图像进行处理,得到最终的重定向结果.基于公开数据库在两个不同评价标准下与多种重定向方法的对比表明,本文方法能够最大程度的保留图像的显著部分.
To solve the problem of deformation and distortion that may occur in the current content-aware redirection method,a retargeting method with seam-carving combining saliency and depth information is proposed.Specifically,the GBVS algorithm is used to obtain the saliency information,and the image gradient information is combined with the image depth information acquired by the SIFT matching method to construct a more accurate importance map.Then,according to the energy distribution of the importance map,the original image is processed to obtain the final results.The comparison on the public database with different methods under two evaluation criteria shows that the proposed method can preserve the significant part of the image to the greatest extent.
引文
[1] Rubinstein M,Shamir A,Avidan S.Multi-operator media retargeting[J].ACM Transactions on Graphics,2009,28(3):341-352.
[2] Wolf L,Guttmann M,Cohenor D.Non-homogeneous content-driven video-retargeting [A].Proceedings of the 11th International Conference on Computer Vision [C].Rio de Janeiro,Brazil:IEEE,2007.1-6.
[3] Avidan S,Shamir A.Seam carving for content-aware image resizing[J].ACM Transactions on Graphics,2007,26(3):1-10.DOI:10.1145/1276377.1276390.
[4] Rubinstein M,Shamir A,Avidan S.Improved seam carving for video retargeting[J].ACM Transactions on Graphics,2008,27(3):1-10.DOI:10.1145/1360612.1360615.
[5] Pritch Y,Kav-Venaki E,Peleg S.Shift-map image editing [A].Proceedings of the 12th International Conference on Computer Vision [C].Kyoto,Japan:IEEE,2010.151-158.
[6] Radhakrishna A,Sabine S.Saliency detection for content-aware image resizing [A].Proceedings of the 16th International Conference on Image Processing [C].Cairo,Egypt:IEEE,2009.1005-1008.
[7] Daniel D,Alexandre A,Pierre V.Stream carving:an adaptive seam carving algorithm [A].Proceedings of the 17th International Conference on Image Processing [C].Hong Kong,China:IEEE,2010.901-904.
[8] Shen J,Wang D,Li X.Depth-aware image seam carving[J].IEEE Transactions on Cybernetics,2013,43(5):1453-1461.
[9] Nguyen P B,Beghdadi A,Luong M.Perceptual watermarking using a new just-noticeable-difference model[J].Signal Processing Image Communication,2013,28(10):1506-1525.
[10] Shafieyan F,Karimi N,Mirmahboub B,et al.Image retargeting using depth assisted saliency map[J].Signal Processing Image Communication,2017,50(C):34-43.
[11] Wang Y S,Tai C L,Sorkine O,et al.Optimized scale-and-stretch for image resizing[J].ACM Transactions on Graphics,2008,27(5):1-8.
[12] Philipp K,Manuel L,Alexander H,et al.A system for retargeting of streaming video[J].ACM Transactions on Graphics,2009,28(5):1-10.DOI:10.1145/1618452.1618472.
[13] Panozzo D,Weber O,Sorkine O.Robust image retargeting via axis-aligned deformation[J].Computer Graphics Forum,2012,31(2pt1):229-236.
[14] Das B C,Gopalakrishnan V,Iyer K N,et al.Similarity and rigidity preserving image retargeting [A].Proceedings of the International Conference on Image Processing [C].Phoenix,AZ,USA:IEEE,2016.1584-1588.
[15] Zhang Y,Ngan K N.Fast patch-wise image retargeting [A].Proceedings of the International Conference on Image Processing [C].Phoenix,AZ,USA:IEEE,2016.1813-1817.
[16] Sch?lkopf B,Platt J,Hofmann T.Graph-based visual saliency[J].Proc of Neural Information Processing Systems,2006,19:545-552.
[17] Liu C,Yuen J,Torralba A,et al.SIFT flow:dense correspondence across different scenes [A].Proceedings of the 10th European Conference on Computer Vision [C].Marseille,France:Springer Verlag,2008.28-42.
[18] Rubinstein M,et al.A comparative study of image retargeting[J].ACM Transactions on Graphics,2010,29(6):160.
[19] Oliveira S A F,Alves S S A,Gomes J P P,et al.A bi-directional evaluation-based approach for image retargeting quality assessment[J].Computer Vision and Image Understanding,2018,168:172-181.