基于CT数据的二次曲面拟合算法研究

樊知轩; 阙介民; 魏存峰; 刘宝东; 魏彪

doi:10.15953/j.ctta.2022.040

基于CT数据的二次曲面拟合算法研究

樊知轩^{1a, 2,},
阙介民²,
魏存峰^{2, 3},
刘宝东^{2, 3, ,},
魏彪^1b, ,

1.
自动化学院；重庆大学
2.
光电工程学院，重庆400044 重庆大学
2.
中国科学院高能物理研究所（北京市射线成像技术与装备工程技术研究中心）, 北京100049
3.
中国科学院大学核科学与技术学院, 北京100049

基金项目: 中国科学院科研仪器设备研制项目（固液界面可视可控的光学晶体生长设备的研制（E12821V1））

详细信息

作者简介:
樊知轩: 男，重庆大学控制工程专业在读硕士研究生，主要从事X射线成像物理、CT成像数学算法及工程应用等方面的研究，E-mail：124356321@qq.com

刘宝东: 男，博士，中国科学院高能物理研究所副研究员，主要从事X射线成像物理、CT成像数学算法及工程应用等方面的研究，E-mail：liubd@ihep.ac.cn

魏彪: 男，博士，重庆大学光电工程学院教授、博士生导师，重庆大学ICT无损检测教育部工程研究中心副主任，主要从事X射线成像物理、CT成像数学算法及工程应用等方面的研究，E-mail：weibiao@cqu.edu.cn

通讯作者:
刘宝东: 男，博士，中国科学院高能物理研究所副研究员，主要从事X射线成像物理、CT成像数学算法及工程应用等方面的研究，E-mail：liubd@ihep.ac.cn

魏彪: 男，博士，重庆大学光电工程学院教授、博士生导师，重庆大学ICT无损检测教育部工程研究中心副主任，主要从事X射线成像物理、CT成像数学算法及工程应用等方面的研究，E-mail：weibiao@cqu.edu.cn

中图分类号: O 242; TP 391.41
计量
- 文章访问数: 366
- HTML全文浏览量: 156
- PDF下载量: 53
出版历程
- 收稿日期: 2022-03-08
- 修回日期: 2022-04-02
- 录用日期: 2022-04-11
- 网络出版日期: 2022-05-30
- 发布日期: 2023-01-30

Research on Quadric Surface Fitting Algorithm Based on CT Data

FAN Zhixuan^{1a, 2,},
QUE Jiemin²,
WEI Cunfeng^{2, 3},
LIU Baodong^{2, 3, ,},
WEI Biao^1b, ,

1.
Chongqing University, Chongqing 400044, China College of Automation
2.
Chongqing University, Chongqing 400044, China College of Optoelectronic Engineering
2.
Institute of High Energy Physics, Chinese Academy of Sciences (Beijing Engineering Research Center of Radiographic Techniques and Equipment), Beijing 100049, China
3.
School of Nuclear Science and Technology, University of Chinese Academy of Sciences, Beijing 100049, China

摘要

摘要: 二次曲面工件在工业中比较常见，为测量物体内部的二次曲面，本文采用工业计算机断层成像（CT）技术获取物体的断层图像序列，利用U-net图像分割网络获得断层图像上的目标区域，对分割结果的边缘进行检测和曲线拟合并堆叠成三维点集，通过曲面拟合获取物体曲面的三维空间坐标信息。研究结果表明，本文的方法能够有效地实现边界提取和界面参数的拟合工作，拟合误差在1％以内，相比传统方法有较大改进。
- 工业CT /
- 深度学习 /
- 二次曲面 /
- 曲面拟合
Abstract: Quadric surface is a common type of workpiece shape in industry. It can be imaged by computed tomography (CT). A sample object was scanned and its slices were reconstructed to measure its internal quadric surface. We used the U-net image segmentation network to obtain the interested region and then detected the edge in the segmentation results and obtained curve fitting results. The curves were stacked into a three-dimensional point set. The three-dimensional spatial coordinate information for the internal quadric surface was computed through surface fitting. The results demonstrate that the proposed method can effectively extract the internal quadric surface parameters, and the fitting error can be controlled within 1%, which is superior to that achieved with a traditional algorithm.
- ICT /
- deep learning /
- quadric surface /
- curving fitting

HTML全文

图 1 模体示意图

Figure 1. Schematic of the model

下载: 全尺寸图片幻灯片

图 2 拟合流程图

Figure 2. Fitting process

下载: 全尺寸图片幻灯片

图 3 U-net网络结构图

Figure 3. U-net network structure diagram

下载: 全尺寸图片幻灯片

图 4 数据预处理流程图

Figure 4. Data preprocessing process

下载: 全尺寸图片幻灯片

图 5 U-net分割结果图

Figure 5. U-net segmentation result

下载: 全尺寸图片幻灯片

图 6 区域生长分割结果图

Figure 6. Region growing segmentation results

下载: 全尺寸图片幻灯片

图 7 边缘检测结果图

Figure 7. Edge detection result graphs

下载: 全尺寸图片幻灯片

图 8 椭圆拟合结果图

Figure 8. Ellipse fitting result diagrams

下载: 全尺寸图片幻灯片

图 9 三维点集图

Figure 9. 3D point set diagram

下载: 全尺寸图片幻灯片

表 1 拟合结果

Table 1 Fitting results

数据精简方式	曲面高度Δh/mm			曲面曲率δ/％
数据精简方式	模体值	拟合值	相对误差/％	模体值	拟合值	相对误差/％
进行异常数据剔除	14.6	14.67	0.40	0.0973	0.0978	0.40
不进行不变层去除	14.6	15.04	3.01	0.0973	0.1002	3.01
不进行稳定层去除	14.6	13.39	8.29	0.0973	0.0892	8.29

下载: 导出CSV

参考文献(20)

[1]	SENGUTTUVAN N, AOSHIMA M, SUMIYA K, et al. Oriented growth of large size calcium Fluoride single crystals for optical lithography[J]. Journal of Crystal Growth, 2005, 280(3/4): 462−466.
[2]	TANDJAOUI A, MANGELINCK-NOEL N, REINHART G, et al. Investigation of grain boundary grooves at the solid-liquid interface during directional solidification of multi-crystalline silicon: In situ characterization by X-ray imaging[J]. Journal of Crystal Growth, 2013, 377(15): 203−211.
[3]	DONG Y, SHUAI S, ZHENG T, et al. In-situ observation of solid-liquid interface transition during directional solidification of Al-Zn alloy via X-ray imaging[J]. Journal of Materials Science and Technology, 2020, 39: 113−123. doi: 10.1016/j.jmst.2019.06.026
[4]	陈福泉, 林樽达, 余炳和. 盐封下降法生长KRS-5晶体[J]. 激光与红外, 1981,(9): 50−56. CHEN F Q, LIN Z D, YU B H. Growth of KRS-5 crystal by salt sealing down method[J]. Laser & Infrared, 1981, (9): 50−56. (in Chinese).
[5]	李晓辉. 坩埚下降法CaF_2基晶体生长的数值模拟及缺陷研究[D]. 北京: 中国科学院大学, 2020. LI X H. Numerical simulation and defect analysis for the growth of alkaline-earth fluorides[D]. Beijing: University of Chinese Academy of Sciences, 2020. (in Chinese).
[6]	曹冰, 何朝明, 李柏林. 工业CT图像轮廓数据提取研究[J]. 机电工程技术, 2009,(9): 3. CAO B, HE C M, LI B L. Extraction of contour data based on industrial CT image[J]. Mechanical & Electrical Engineering Technology, 2009, (9): 3. (in Chinese).
[7]	LUKÁCS G, MARTIN R R, MARSHALL A D. Faithful least-squares fitting of spheres, cylinders, cones and tori for reliable segmentation[C]//Computer Vision - ECCV'98, 5th European Conference on Computer Vision, Freiburg, Germany, June 2-6, 1998, Proceedings, Volume I. Springer Berlin Heidelberg, 1998.
[8]	邹斌, 曾理, 马睿. 工业CT图像管道拟合[J]. 计算机工程与应用, 2010,46(11): 208−210. doi: 10.3778/j.issn.1002-8331.2010.11.063 ZOU B, ZENG L, MA R. Fiting of industrial CT image pipes[J]. Computer Engineering and Applications, 2010, 46(11): 208−210. (in Chinese). doi: 10.3778/j.issn.1002-8331.2010.11.063
[9]	卢艳平, 王珏, 覃仁超. 一种剥皮算法在工业CT图像分割中的应用[J]. 无损检测, 2005,27(5): 235−237. doi: 10.3969/j.issn.1000-6656.2005.05.004 LU Y P, WANG J, QIN R C. Application of peeling method to segmentation of industrial computed tomographic image[J]. Nondestructive Testing, 2005, 27(5): 235−237. (in Chinese). doi: 10.3969/j.issn.1000-6656.2005.05.004
[10]	KOHLER R. A segmentation system based on thresholding[J]. Computer Graphics and Image Processing, 1981, 15(4): 319−338. doi: 10.1016/S0146-664X(81)80015-9
[11]	ADAMS R, BISHOF L. Seeded region growing[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1994, 16(6): 641-647.
[12]	SHEN X, SPANN M. Segmentation of 2D and 3D images through a hierarchical clusteringbased on region modelling[C]//Image Processing, 1997. Proceedings. International Conference on. IEEE, 1997.
[13]	SHELHAMER E, JONATHAN L, DARRELL T. Fully convolutional networks for semantic segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(4): 640−651.
[14]	RONNEBERGER O, FISCHER P, BROX T. U-net: Convolutional networks for biomedical image segmentation[J]. Springer International Publishing, 2015.
[15]	ZHAO H, SHI J, QI X, et al. Pyramid scene parsing network[J]. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Honolulu: IEEE, 2017, 6230-6239.
[16]	CHEN L C, PAPANDREOU G, KOKKINOS I, et al. Semantic image segmentation with deep convolutional nets and fully connected CRFs[C]//Computer Science. 2014: 357-361.
[17]	CHEN L C, PAPANDREOU G, KOKKINOS I, et al. DeepLab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40(4): 834−848. doi: 10.1109/TPAMI.2017.2699184
[18]	CHEN L C, PAPANDREOU G, SCHROFF F, et al. Rethinking atrous convolution for semantic image segmentation[J]. arXiv preprint arXiv: 1706.05587, 2017.
[19]	FITZGIBBON A, PILU M, FISHER R B. Direct least square fitting of ellipses[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 1999.
[20]	MICELI A, THIERRY R, BETTUZZI M, et al. Comparison of simulated and measured spectra of an industrial 450 kV X-ray tube[J]. Nuclear Instruments & Methods in Physics Research. Section A, Accelerators, Spectrometers, Detectors and Associated Equipment, 2007, 580(1): 123-126.