多特征混合的流形排序及其在三维模型草图检索中的应用
|
摘要
为了解决现有基于流形排序的三维模型的草图检索方法特征提取过程中特征描述不准确,且需要对检索草图进行人工标注的问题,提出了一种基于改进特征描述符和深度学习的方法用于三维模型草图检索。该方法首先综合考虑了整体形状及局部细节对检索性能的影响,提出用于描述草图和三维模型投影视图的多特征视觉描述符。然后利用深度学习的方法实现草图语义标注。最后在包含7 200幅草图和1 258个三维模型的公开数据集上进行了实验验证。结果表明:本文方法不仅降低了人为标注所带来的干扰,而且显著提高了三维模型检索的准确率。研究结果将为三维影视动画的自动化检索及编辑重用等相关应用提供设计思路与技术支撑。
In order to solve the problems of the feature description is not accurate in the feature extraction process and the need to manually mark the search sketch in the existing 3 D model sketch retrieval method based on manifold ranking,a method based on improved feature descriptor and deep learning for 3 D model sketch retrieval.was proposed. Firstly,considering the influence of the overall shape and local details on the retrieval performance,a multi-features visual descriptor is proposed. Then,the input sketch is automatically semantically annotated by the deep learning method. Finally,experiments were carried out on an open dataset containing 7 200 sketches and1 258 3 D models. The results show that the proposed method not only reduces the interference caused by human annotation,but also significantly improves the accuracy of 3 D model retrieval. The research results will provide design ideas and technical support for related applications such as automated retrieval and editing reuse of 3 D film and television animation.
In order to solve the problems of the feature description is not accurate in the feature extraction process and the need to manually mark the search sketch in the existing 3 D model sketch retrieval method based on manifold ranking,a method based on improved feature descriptor and deep learning for 3 D model sketch retrieval.was proposed. Firstly,considering the influence of the overall shape and local details on the retrieval performance,a multi-features visual descriptor is proposed. Then,the input sketch is automatically semantically annotated by the deep learning method. Finally,experiments were carried out on an open dataset containing 7 200 sketches and1 258 3 D models. The results show that the proposed method not only reduces the interference caused by human annotation,but also significantly improves the accuracy of 3 D model retrieval. The research results will provide design ideas and technical support for related applications such as automated retrieval and editing reuse of 3 D film and television animation.
引文
1段竹.三维模型检索技术综述[J].科技资讯,2006(6):4-5Duan Zhu. A survey of 3D model retrieval techniques[J]. Science&Technology Information,2006(6):4-5
2 牟春倩,唐雁,胡金戈.基于流形排序的三维模型检索方法[J].山东大学学报(工学版),2017,47(4):19-24Mou Chunyan,Tang Yan,Hu Jinge. A new 3D model retrieval method based on manifold ranking[J]. Journal of Shandong University(Engineering Science),2017,47(4):19-24
3 李海生,孙莉,武玉娟,等.非刚性三维模型检索特征提取技术研究[J].软件学报,2018,29(2):483-505Li Haisheng,Sun Li,Wu Yujuan,et al. Survey on feature extraction techniques for non-rigid 3D shape retrieval[J]. Journal of Software,2018,29(2):483-505
4 解亮.基于最优草图局部多尺度GABOR特征的3D模型检索[D].长春:吉林大学,2016Xie Liang. 3D model retrieving based on local multi-scale GABOR feature of best sketch[D]. Changchun:Jilin University,2016
5 钱露,樊亚春,周明全,等.基于Leap Motion的三维模型草图检索[J].图学学报,2017,38(2):267-277Qian Lu,Fan Yachun,Zhou Mingquan,et al. Sketch-based threedimensional model retrieval based on leap motion[J]. Journal of Graphics,2017,38(2):267-277
6 Li B,Johan H. Sketch-based 3D model retrieval by incorporating 2D-3 D alignment[J]. Multimedia Tools Applications,2013,65(3):363-385
7 Fonseca M J,Ferreira A,Jorge J. Sketch-based interfaces and modeling[M]. London:Springer,2011:181-201
8 Sousa P,Fonseca M J. Sketch-based retrieval of drawings using spatial proximity[J]. Journal of Visual Languages&Computing,2010,21(2):69-80
9 Saavedra J M,Bustos B. An improved histogram of edge local orientations for sketch-based image retrieval[M]. Heidelberg:Springer,2010:432-441
10 Furuya T, Ohbuchi R. Ranking on cross-domain manifold for sketch-based 3D model retrieval[C]//International Conference on Cyberworlds. Washington,D. C.:IEEE Computer Society,2013:274-281
11 Eitz M,Hildebrand K,Boubekeur T,et al. Sketch-based 3D shape retrieval[C]//ACM Siggraph. New York:ACM,2010:5
12 Hu M. Visual pattern recognition by moment invariants[J]. Ire Transactions on Information Theory,1962,8(2):179-187
13 Lowe D G. Distinctive image features from scale-invariant key points[J]. International Journal of Computer Vision,2004,60(2):91-110
14 Li B,Schreck T,Godil A,et al. SHREC'12 track:Sketch-based3 D shape retrieval[C]//Proceedings of the 5th Eurographics conference on 3D Object Retrieval. Aire-la-Ville:Eurographics Association,2012:109-118
15 孙志军,薛磊,许阳明,等.深度学习研究综述[J].计算机应用研究,2012,29(8):2806-2810Sun Zhijum,Xue Lei,Xu Yangming,et al. Overview of deep learning[J]. Application Research of Computers,2012,29(8):2806-2810
16 Zhou D,Bousquet O,Lal T N,et al. Learning with local and global consistency[J]. Advances in Neural Information Processing Systems,2004,16(4):321-328.
17 Li B,Lu Y,Godil A,et al. SHREC'13 track:Large scale sketchbased 3D shape retrieval[C]//Proceedings of the Sixth Eurographics Workshop on 3D Object Retrieval. Aire-la-Ville:Eurographics Association,2013:89-96
18 Shilane P,Min P,Kazhdan M,et al. The princeton shape benchmark[C]//Proceedings of the International Conference on Shape Modeling and Applications. New York:IEEE Press,2004:167-178
19 Aono M,Iwabuchi H. 3D shape retrieval from a 2D image as query[C]//Signal&Information Processing Association Summit and Conference. New York:IEEE,2013:1-10
2 牟春倩,唐雁,胡金戈.基于流形排序的三维模型检索方法[J].山东大学学报(工学版),2017,47(4):19-24Mou Chunyan,Tang Yan,Hu Jinge. A new 3D model retrieval method based on manifold ranking[J]. Journal of Shandong University(Engineering Science),2017,47(4):19-24
3 李海生,孙莉,武玉娟,等.非刚性三维模型检索特征提取技术研究[J].软件学报,2018,29(2):483-505Li Haisheng,Sun Li,Wu Yujuan,et al. Survey on feature extraction techniques for non-rigid 3D shape retrieval[J]. Journal of Software,2018,29(2):483-505
4 解亮.基于最优草图局部多尺度GABOR特征的3D模型检索[D].长春:吉林大学,2016Xie Liang. 3D model retrieving based on local multi-scale GABOR feature of best sketch[D]. Changchun:Jilin University,2016
5 钱露,樊亚春,周明全,等.基于Leap Motion的三维模型草图检索[J].图学学报,2017,38(2):267-277Qian Lu,Fan Yachun,Zhou Mingquan,et al. Sketch-based threedimensional model retrieval based on leap motion[J]. Journal of Graphics,2017,38(2):267-277
6 Li B,Johan H. Sketch-based 3D model retrieval by incorporating 2D-3 D alignment[J]. Multimedia Tools Applications,2013,65(3):363-385
7 Fonseca M J,Ferreira A,Jorge J. Sketch-based interfaces and modeling[M]. London:Springer,2011:181-201
8 Sousa P,Fonseca M J. Sketch-based retrieval of drawings using spatial proximity[J]. Journal of Visual Languages&Computing,2010,21(2):69-80
9 Saavedra J M,Bustos B. An improved histogram of edge local orientations for sketch-based image retrieval[M]. Heidelberg:Springer,2010:432-441
10 Furuya T, Ohbuchi R. Ranking on cross-domain manifold for sketch-based 3D model retrieval[C]//International Conference on Cyberworlds. Washington,D. C.:IEEE Computer Society,2013:274-281
11 Eitz M,Hildebrand K,Boubekeur T,et al. Sketch-based 3D shape retrieval[C]//ACM Siggraph. New York:ACM,2010:5
12 Hu M. Visual pattern recognition by moment invariants[J]. Ire Transactions on Information Theory,1962,8(2):179-187
13 Lowe D G. Distinctive image features from scale-invariant key points[J]. International Journal of Computer Vision,2004,60(2):91-110
14 Li B,Schreck T,Godil A,et al. SHREC'12 track:Sketch-based3 D shape retrieval[C]//Proceedings of the 5th Eurographics conference on 3D Object Retrieval. Aire-la-Ville:Eurographics Association,2012:109-118
15 孙志军,薛磊,许阳明,等.深度学习研究综述[J].计算机应用研究,2012,29(8):2806-2810Sun Zhijum,Xue Lei,Xu Yangming,et al. Overview of deep learning[J]. Application Research of Computers,2012,29(8):2806-2810
16 Zhou D,Bousquet O,Lal T N,et al. Learning with local and global consistency[J]. Advances in Neural Information Processing Systems,2004,16(4):321-328.
17 Li B,Lu Y,Godil A,et al. SHREC'13 track:Large scale sketchbased 3D shape retrieval[C]//Proceedings of the Sixth Eurographics Workshop on 3D Object Retrieval. Aire-la-Ville:Eurographics Association,2013:89-96
18 Shilane P,Min P,Kazhdan M,et al. The princeton shape benchmark[C]//Proceedings of the International Conference on Shape Modeling and Applications. New York:IEEE Press,2004:167-178
19 Aono M,Iwabuchi H. 3D shape retrieval from a 2D image as query[C]//Signal&Information Processing Association Summit and Conference. New York:IEEE,2013:1-10
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |