水下目标三维点云生成优化方法
Optimization Method of 3D Point Cloud Generation for Underwater Target
DOI: 10.12677/OJTT.2022.111004, PDF,    国家科技经费支持
作者: 袁利毫, 杨永俊, 昝英飞, 秦 浩：哈尔滨工程大学船舶工程学院，黑龙江 哈尔滨
关键词: 特征点检测；图像匹配；三维重构；点云模型；Feature Point Detection； Image Matching； 3D Reconstruction； Point Cloud Model
摘要: 由于水下光衰严重、能见度低，导致ROV作业时采集的水下图像特征难以提取，重建的水下目标三维点云辨识质量低。本文提出一种改进SIFT尺度空间的水下目标三维点云生成优化方法，该方法增强了特征提取时颜色边缘特征信息的感知，提高了有效特征点匹配数。基于ROV拍摄的水下数据集实验结果表明，相比于SIFT传统算法，该方法图像特征点数量增加2.9倍，匹配数量增加1.78倍，点云数量增加1.04倍，三维重构模型效果显著。因此该方法可在一定程度解决水下图像特征点难检测及生成的三维模型质量差的问题。
Abstract: Due to the serious underwater light attenuation and low visibility, it is difficult to extract the un-derwater image features collected during ROV operation, and the recognition quality of the reconstructed underwater target 3D point cloud is low. In this paper, an optimization method for underwater target 3D point cloud generation based on improved SIFT scale space is proposed. This method enhances the perception of color edge feature information and improves the number of effective feature points. The experimental results of underwater dataset taken based on ROV show that compared with the traditional SIFT algorithm, the number of image feature points, matching and point clouds are increased by 2.9 times, 1.78 times and 1.04 times respectively. The effect of 3D reconstruction model is remarkable. Therefore, this method can solve the problem of difficult detection of underwater image feature points and poor quality of 3D model to a certain extent.
文章引用：袁利毫, 杨永俊, 昝英飞, 秦浩. 水下目标三维点云生成优化方法[J]. 交通技术, 2022, 11(1): 31-43. https://doi.org/10.12677/OJTT.2022.111004

参考文献

[1] 张阳, 徐爽, 朱建军, 等. 水下视觉SLAM图像增强研究[J]. 海洋信息, 2020, 35(4): 29-34.
[2] 李海森, 高珏, 徐超, 等. 应用TBD的多波束声呐图像序列SIFT特征追踪[J]. 哈尔滨工程大学学报, 2016, 37(8): 1009-1014.
[3] 王昕平, 张森林, 刘妹琴, 樊臻. 基于多尺度图像融合和SIFT特征的水下图像拼接研究[J]. 计算机应用与软件, 2021, 38(5): 213-217+230.
[4] Lucas, B. (1981) An Iterative Image Registration Technique with an Application to Stereo Vision (DARPA). Proceeding IJCAI, 81, 674-679.
[5] Tomasi, C. and Kanade, T. (1991) Detection and Tracking of Point Features. International Journal of Computer Vision, 9, 137-154. [Google Scholar] [CrossRef] 
[6] Birchfield, S. (1997) Derivation of Kanade-Lucas-Tomasi Tracking Equation. Unpublished Notes, 44, 1811-1843.
[7] Frahm, S., Pollefeys, M. and Genc, Y. (2000) GPU-Based Video Feature Tracking and Matching. Workshop on Edge Computing Using New Commodity Architectures, 2000, 189-196.
[8] Sinha, S.N., Frahm, J.M., Pollefeys, M., et al. (2011) Feature Tracking and Matching in Video Using Programmable Graphics Hardware. Machine Vision Applications, 22, 207-217. [Google Scholar] [CrossRef] 
[9] Poling, B., Lerman, G. and Szlam, A. (2014) Better Feature Tracking through Subspace Constraints. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Columbus, 23-28 June 2014, 3454-3461. [Google Scholar] [CrossRef] 
[10] Hwangbo, M., Kim, J.S. and Kanade, T. (2009) Inertial-Aided KLT Feature Tracking for a Moving Camera. Proceedings of the 2009 IEEE/RSJ International Conference on Intelligent Robots and Systems, St Louis, 11-15 October 2009, 1909-1916. [Google Scholar] [CrossRef] 
[11] Wilson, K. and Snavely, N. (2013) Network Principles for SfM: Disambiguating Repeated Structures with Local Context. Proceedings of the IEEE International Conference on Computer Vision, Sydney, 1-8 December 2013, 513-520. [Google Scholar] [CrossRef] 
[12] Ceylan, D., et al. (2014) Coupled Structure-from-Motion and 3D Symmetry Detection for Urban Facades. ACM Transactions on Graphics, 33, 1-15. [Google Scholar] [CrossRef] 
[13] Endres, F., Hess, J., Engelhard, N., et al. (2012) An Evaluation of the RGB-D SLAM System. Proceedings IEEE International Conference on Robotics Automation, St. Paul, 14-18 May 2012. [Google Scholar] [CrossRef] 
[14] Engel, J., Schps, T. and Cremers, D. (2014) LSD-SLAM: Large-Scale Direct Monocular SLAM. 13th European Conference on Computer Vision—ECCV, Zurich, 6-12 September 2014, 834-849. [Google Scholar] [CrossRef] 
[15] Zhang, G., Dong, Z., Jia, J., et al. (2010) Efficient Non-Consecutive Feature Tracking for Structure-from-Motion. European Conference on Computer Vision, Heraklion, 5-11 September 2010, 422. [Google Scholar] [CrossRef] 
[16] Zhang, G. and Vela, P.A. (2015) Good Features to Track for Visual SLAM. 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Boston, 7-12 June 2015, 1373-1382. [Google Scholar] [CrossRef] 
[17] Garrigues, M. and Manzanera, A. (2012) Real Time Semi-Dense Point Tracking. Proceedings of the International Conference on Image Analysis and Recognition (ICIAR 2012), Aveiro, 25-27 June 2012, 245-252.
[18] Lee, T. and Hollerer, T. (2008) Hybrid Feature Tracking and User Interaction for Markerless Augmented Reality. Proceedings of the Virtual Reality Conference, 2008 VR ‘08 IEEE, Reno, 8-12 March 2008, 145-152. [Google Scholar] [CrossRef] 
[19] Buchanan, A. and Fitzgibbon, A. (2006) Interactive Feature Tracking Using K-D Trees and Dynamic Programming. 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR’06), New York, 17-22 June 2006, 626-633.
[20] Zhang, G., Liu, H., Dong, Z., et al. (2015) ENFT: Efficient Non-Consecutive Feature Tracking for Robust Structure-from-Motion. IEEE Transactions on Image Processing, 25, 5957-5970. [Google Scholar] [CrossRef] 
[21] Wu, C., Clipp, B., Li, X., et al. (2008) 3D Model Matching with Viewpoint-Invariant Patches (VIP). 2008 IEEE Conference on Computer Vision and Pattern Recognition, Anchorage, 23-28 June 2008, 1-8.
[22] Zach, C., Klopschitz, M. and Pollefeys, M. (2010) Disambiguating Visual Relations Using Loop Constraints. 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, San Francisco, 13-18 June 2010, 1426-1433. [Google Scholar] [CrossRef] 
[23] Trees, G.H. (2012) Point Track Creation in Unordered Image Collections Using Gomory-Hu Trees. Proceedings of the 21st International Conference on Pattern Recognition (ICPR2012), Tsukuba, 11-15 November 2012, 2116-2119.
[24] Gomory, R.E. and Hu, T.C. (1961) Multi-Terminal Network Flows. Journal of SIAM, 9, 551-570. [Google Scholar] [CrossRef] 

为你推荐