[1]陈起凤,刘 军*,李 威,等.三维重建中线段匹配方法的研究[J].武汉工程大学学报,2018,40(04):446-450.[doi:10. 3969/j. issn. 1674?2869. 2018. 04. 019]
CHEN Qifeng,LIU Jun*,LI Wei,et al. Line Matching in Three-Dimensional Reconstruction[J].Journal of Wuhan Institute of Technology,2018,40(04):446-450.[doi:10. 3969/j. issn. 1674?2869. 2018. 04. 019]
点击复制
《武汉工程大学学报》[ISSN:1674-2869/CN:42-1779/TQ]
- 卷:
-
40
- 期数:
-
2018年04期
- 页码:
-
446-450
- 栏目:
-
机电与信息工程
- 出版日期:
-
2018-08-23
文章信息/Info
- Title:
-
Line Matching in Three-Dimensional Reconstruction
- 文章编号:
-
20180419
- 作者:
-
陈起凤1; 2; 刘 军*1; 2; 李 威1; 2; 雷光元3; 董广峰3
-
1. 智能机器人湖北省重点实验室(武汉工程大学),湖北 武汉 430205;2. 武汉工程大学计算机科学与工程学院,湖北 武汉 430205;3. 国投新疆罗布泊钾盐有限责任公司,新疆 哈密 893000
- Author(s):
-
CHEN Qifeng1; 2; LIU Jun*1; 2; LI Wei1; 2; LEI Guangyuan3; DONG Guangfeng3
-
1. Hubei Key Laboratory of Intelligent Robot (Wuhan Institute of Technology), Wuhan 430205, China;2. School of Computer Science & Engineering, Wuhan Institute of Technology, Wuhan 430205, China;3. SDIC Xinjiang Luobupo Potash Co., LTD, Hami 839000, China
-
- 关键词:
-
三维重建; 线段匹配; 直线提取算法; 参数评估
- Keywords:
-
3D reconstruction; line matching; line segment detector algorithm; parameter evaluation
- 分类号:
-
TP317.4
- DOI:
-
10. 3969/j. issn. 1674?2869. 2018. 04. 019
- 文献标志码:
-
A
- 摘要:
-
针对传统基于特征点匹配的重建模型容易存在空洞现象的问题,提出一种基于直线提取算法的线段匹配方法。首先使用无人机设备获取校园建筑图片集,保证拍摄过程中相邻图片存在重叠部分,通过筛选分数最高的相似性评分得到相邻图片集合并只在相邻图片间进行匹配。接着利用直线提取算法提取图片中直线段,然后根据极线约束关系进行二维线段匹配最终生成三维线段模型。二维线段匹配阈值[τ]、相邻图片个数M、在相邻图像中与特定二维线段相匹配的线段个数K是其中3个重要参数,对4组不同的图片集进行参数评估,实验结果表明当二维线段阈值[τ]低于0.5,相邻图片个数M为10,在相邻图像中与特定二维线段相匹配的线段个数K为10时三维重建速率与模型精度可接受度最高。
- Abstract:
-
A line matching method based on Line Segment Detector (LSD) algorithm was proposed to resolve holes possibly in traditional three-dimensional(3D) reconstruction by characteristic point matching. Firstly, we used Unmanned Aerial Vehicle to take pictures of campus buildings and ensured that there were overlaps in neighboring pictures. Secondly, we selected the highest similarity scores to obtain the neighboring picture sets and matched only between neighboring pictures. Then, LSD algorithm was used to extract the line segments in the pictures. Finally, the 3D segment model was reconstructed by using two-dimensional segment matching based on epipolar constraint. Three important parameters, fixed threshold [τ], number of the nearest neighbors M and matched specific segment in neighbors K were used in the model. A parameter evaluation was made in four different picture sets. The experiment results showed that the reconstructing rate and model accuracy were highest when [τ] was less than 0.5, M was 10 and K was 10.
参考文献/References:
[1] 王传安,葛华,王亚军. 基于非完全拓扑分析的三维建筑识别算法[J]. 武汉工程大学学报, 2011, 33(10):100-102. [2] 胡红权. 视频监控中运动图像序列三维重建[J]. 现代电子技术, 2018,41(3):67-70. [3] 徐超,李乔. 基于计算机视觉的三维重建技术综述[J]. 数字技术与应用,2017(1):54-56. [4] 刘军,李娜,刘鹏. 双目视觉立体标定方法的改进[J]. 武汉工程大学学报,2013,35(10):68-73. [5] 郭盛威,章秀华,范艳,等. 三维重建表面几何特征的提取与参数测量计算[J]. 武汉工程大学学报,2016,38(2):186-188. [6] 章秀华,白浩玉,李毅. 多目立体视觉三维重建系统的设计[J]. 武汉工程大学学报,2013,35(3):70-74. [7] 肖永强,王海辉,刘奥丽,等. 双目实时目标三维测量实现方法的研究[J]. 武汉工程大学学报,2016, 38(4):386-393. [8] 杨辉,刘军,阮松. 基于Mean Shift 算法视频跟踪研究[J]. 计算机工程与设计,2013,34(6):2062-2066. [9] 张琳彦,朱利等. 以多幅图像非几何约束线段匹配重建建筑物外立面三维线段模型[J]. 西安交通大学学报,2014,48(4):16-25. [10] 翟优?,郭希维,何鹏,?等. 利用共线性和对极约束实现匹配扩展[J]. 光学技术,2017(5):450-454. [11] 欧元汉,?黄鸿,?陆宇海. 基于密集点云重投影的三维线段模型重建算法[J]. 计算机工程与设计, 2015(12):3291-3296. [12] LOURAKIS?M?I?A,?TZURBAKIS?S?V,ARGYROS A A,et al.? Feature? transfer? and? matching? in? disparate? stereo? views ?through ?the use?of plane homographies[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2003, 25(2):271-276. [13] BAY H, FERRARI V, GOOL L V. Wide-baseline stereo matching with line segments[C] //IEEE Computer Society Conference on Computer Vision and Pattern . IEEE, 2005:329-336 vol. 1. [14] BAY H, TUYTELAARS T, GOOL L V. SURF: speeded up robust features[J]. Computer Vision & Image Understanding, 2006, 110(3):404-417. [15] WANG Z, LIU H, WU F. MSLD: a robust descriptor for line matching[J]. Pattern Recognition, 2009, 42(5):941-953. [16] KIM H, LEE S. A novel line matching method based on intersection context[C] // IEEE International Conference on Robotics and Automation. IEEE, 2010:1014-1021. [17] GROMPONE R,JAKUBOWICZ J,MOREL J M,et al.LSD: a fast line segment detector with afalse detection control[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2010,32(4):722-732. [18] HOCQUELET A, TRILLAUD H, FRULIO N, et al. Three-dimensional measurement of hepatocellular carcinoma ablation zones and margins for predicting local tumor progression[J]. Journal of Vascular & Interventional Radiology Jvir,2016,27(7):1038-1045.
备注/Memo
- 备注/Memo:
-
收稿日期:2018-03-16基金项目:智能机器人湖北省重点实验室开放基金(HBIR201608);武汉工程大学第九届研究生教育创新基金(CX2017076)作者简介:陈起凤,硕士研究生。E-mail:[email protected]*通讯作者:刘 军,博士,副教授。E-mail:[email protected]引文格式:陈起凤,刘军,李威,等. 三维重建中线段匹配方法的研究[J]. 武汉工程大学学报,2018,40(4):446-450.
更新日期/Last Update:
2018-08-16