全部 标题 作者
关键词 摘要

OALib Journal期刊
ISSN: 2333-9721
费用:99美元

查看量下载量

相关文章

更多...

基于神经网络的由单幅二维工程图形重建三维形体系统的研究

Keywords: 神经网络,三维重建,特征提取,OpenInventor

Full-Text   Cite this paper   Add to My Lib

Abstract:

针对基于单幅图形的三维重建方法的多解性和病态性的难点问题,设计实现了一套三维重建系统,提出一种基于角点特征和图形结构特征的人工神经网络分类算法,对输入图形进行识别分类,然后分别对识别结果精确重建,从而避免了直接恢复深度信息的病态解问题.为有效地提取角点特征,采用改进的基于曲率的自适应角点检测算法,给出了滑动窗口的自适应调整策略,并得到图像边界点的局部支持域的特征向量,从而使提取的角点曲率特征具有旋转、平移和尺度不变性.实验验证了改进算法的快速、准确和稳定的特性.

References

[1]  MARSTON R E,KUO M H.Reconstruction of 3D objects from three orthographic projections using decision chainingmethod[C]∥Proceeding of IAPR Workshop on Machine Vision Applications.Kawasaki:[s.n.],1994:423-426.
[2]  SEITZ S M,DYER C R.Photorealistic scene reconstruction by voxel coloring,proc[C]∥Computer Vision and PatternRecognition Conf.Los Alamitos,USA:IEEE Comput Soc,1997:1067-1073.
[3]  SCHARSTEIN D,SZELISKI R.A txaonomy and evaluation of dense two-frame stereo correspondence algorithms[C]∥Proceedings IEEE Workshop on Stereo and Multi Baseline Vision.Los Alamitos,USA:IEEE Comput Soc,2001:131-140.
[4]  ZELEZNIK R C,HERNDON K P,HUGHES J F.Sketch:an interface for sketching 3D scenes[C]∥Proceedings ofSIGGRAPH’96 Conference.New York,USA:ACM,1996:161-170.
[5]  EGGLI L,HSU C,BR DERLIN B D,et al.Inferring 3D models from freehand sketches and constraints[J].Computer-Aided Design,1997,29(2):101-112.
[6]  IGARASHI T,MATSUOKA S,TANAKA H.Teddy:a sketching interface for 3D freeform design[C]∥ACM SIGGRAPHConference Proceedings.New York,USA:ACM,1999:409-416.
[7]  MOKHTARIAN F,MACKWORTH A.Scale-based description and recognition of planar curves and two-dimensional shapes[J].IEEE Trans.Pattern Recognition and Machine Intelligence.1986,8(1):34-43.
[8]  肖茜,鲁宏伟.基于高斯平滑的自适应角点检测[J].计算机辅助设计与图形学学报,2003,15(11):1358-1361.XIAO Qian,LU Hong-wei.Adaptive corner detection based on gaussian smoothing[J].Journal of Computer Aided Design&Computer Graphics,2003,15(11):1358-1361.
[9]  ROSENFELD A,WESZKA J A.An improved method of angle detection on digital curves[J].IEEE Trans Computers,1975,24:940-941.

Full-Text

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133