点云场景语义标注的排序批处理模式主动学习法

doi:10.11947/j.AGCS.2023.20210332

测绘学报 ›› 2023, Vol. 52 ›› Issue (2): 260-271.doi: 10.11947/j.AGCS.2023.20210332

点云场景语义标注的排序批处理模式主动学习法

邹禄杰^1,2,3, 花向红^1,2, 赵不钒^1,2, 陶武勇^1,2,4, 李琪琪^1,2,3

1. 武汉大学测绘学院, 湖北武汉 430079;
2. 武汉大学灾害监测与防治研究中心, 湖北武汉 430079;
3. 广州市城市规划勘测设计研究院, 广东广州 510060;
4. 南昌大学信息工程学院, 江西南昌 330031

收稿日期:2021-06-28 修回日期:2022-05-18 发布日期:2023-03-07
通讯作者: 花向红 E-mail:xhhua@sgg.whu.edu.cn
作者简介:邹禄杰(1996-),男,硕士生,研究方向为点云数据处理。E-mail:2015301610030@whu.edu.cn
基金资助:
国家自然科学基金（41674005；41871373）

Ranked batch-mode active learning method for semantic annotation of point cloud scene

ZOU Lujie^1,2,3, HUA Xianghong^1,2, ZHAO Bufan^1,2, TAO Wuyong^1,2,4, LI Qiqi^1,2,3

1. School of Geodesy and Geomatics, Wuhan University, Wuhan 430079, China;
2. Research Center for Hazard Monitoring and Prevention, Wuhan University, Wuhan 430079, China;
3. Guangzhou Urban Planning & Design Survey Research Institute, Guangzhou 510060, China;
4. School of Information Engineering, Nanchang University, Nanchang 330031, China

Received:2021-06-28 Revised:2022-05-18 Published:2023-03-07
Supported by:
The National Natural Science Foundation of China (Nos. 41674005;41871373)

摘要/Abstract

摘要： 针对点云场景语义标注存在着手工标注费时费力、算法耗时严重、标注精度不高和不适用于大规模场景点云处理等问题，本文提出了一种结合排序批处理模式的主动学习点云场景语义标注方法。该方法首先对原始点云进行下采样处理，然后利用改进的递归特征增加法从庞大的特征集中筛选出最优特征子集，采用排序批处理模式采样算法迭代选取并人工标注少数未标注点，通过创建最小人工标注训练集来完成下采样点云的语义标注工作，最后利用邻域等权标签传播算法完成原始点云数据的标注。对3个室外大场景点云分别进行的试验表明：本文方法只需人工标注7.50%、7.35%、5.83%的点云即可完成下采样点云的标注工作。此外，对比试验表明，本文方法在标注精度和减少人工成本方面优于其他方法，能为点云语义标注工作节省大量人工成本。

关键词: 点云场景, 排序批处理模式, 特征选择, 主动学习, 语义标注

Abstract: Due to the semantic annotation of point cloud scene by manual annotation is time-consuming and label cost process, and the annotation accuracy is not high, the point cloud processing is not suitable for large-scale scenes, this paper proposed an active learning annotation method of point cloud based on ranked batch-mode. This method firstly downsampling the original point cloud, then an improved recursive feature addition method is used to filter out the optimal feature subset from a huge feature set, and a ranked batch-mode sampling algorithm is adopted to iteratively select and manually label fraction of unlabeled points. The semantic annotation of the down-sampled point cloud is completed by creating a minimum manual annotation training set, and finally the original point cloud data is annotated using the neighborhood equal-weight label propagation algorithm. Experiments on three outdoor large scene point clouds show that the method in this paper only needs to manually label 7.50%, 7.35%, and 5.83% of the point clouds to complete the labeling of the down-sampled point clouds. In addition, comparative experiments show that this method is superior to other methods in labeling accuracy and reducing labor costs, and save a lot of labor costs for point cloud semantic annotation work.

Key words: point cloud scene, ranked batch-mode, feature selection, active learning, semantic annotation

中图分类号:

P237

邹禄杰, 花向红, 赵不钒, 陶武勇, 李琪琪. 点云场景语义标注的排序批处理模式主动学习法[J]. 测绘学报, 2023, 52(2): 260-271.

ZOU Lujie, HUA Xianghong, ZHAO Bufan, TAO Wuyong, LI Qiqi. Ranked batch-mode active learning method for semantic annotation of point cloud scene[J]. Acta Geodaetica et Cartographica Sinica, 2023, 52(2): 260-271.

参考文献

[1] 李德仁, 邵振峰, 于文博, 等. 基于时空位置大数据的公共疫情防控服务让城市更智慧[J]. 武汉大学学报(信息科学版), 2020, 45(4):475-487, 556. LI Deren, SHAO Zhenfeng, YU Wenbo, et al. Public epidemic prevention and control services based on big data of spatio-temporal location make cities more smart[J]. Geomatics and Information Science of Wuhan University, 2020, 45(4):475-487, 556.
[2] 刘经南, 詹骄, 郭迟,等.智能高精地图数据逻辑结构与关键技术[J]. 测绘学报, 2019, 48(8):939-953. DOI:10.11947/j.AGCS.2019.20190125. LIU Jingnan, ZHAN Jiao, GUO Chi, et al. Data logic structure and key technologies on intelligent high-precision map[J]. Acta Geodaetica et Cartographica Sinica, 2019, 48(8):939-953. DOI:10.11947/j.AGCS.2019.20190125.
[3] 李德仁, 姚远, 邵振峰. 智慧城市中的大数据[J]. 武汉大学学报(信息科学版), 2014, 39(6):631-640. LI Deren, YAO Yuan, SHAO Zhenfeng. Big data in smart city[J]. Geomatics and Information Science of Wuhan University, 2014, 39(6):631-640.
[4] 杨必胜, 董震. 点云智能研究进展与趋势[J]. 测绘学报, 2019, 48(12):1575-1585. DOI:10.11947/j.AGCS.2019.20190465. YANG Bisheng, DONG Zhen. Progress and perspective of point cloud intelligence[J]. Acta Geodaetica et Cartographica Sinica, 2019, 48(12):1575-1585. DOI:10.11947/j.AGCS.2019.20190465.
[5] ZHAO Bufan, HUA Xianghong, YU Kegen, et al. Indoor point cloud segmentation using iterative Gaussian mapping and improved model fitting[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 58(11):7890-7907.
[6] ZHANG Liangpei, ZHANG Yun, CHEN Zhenzhong, et al. Splitting and merging based multi-model fitting for point cloud segmentation[J]. Journal of Geodesy and Geoinformation Science, 2020, 2(2):78-89.
[7] 蒋腾平, 王永君, 张林淇, 等.融合CNN和MRF的激光点云层次化语义分割方法[J]. 测绘学报, 2021, 50(2):215-225. DOI:10.11947/j.AGCS.2021.20220095. JIANG Tengping, WANG Yongjun, ZHANG Linqi, et al. A LiDAR point cloud hierarchical semantic segmentation method combining CNN and MRF[J]. Acta Geodaetica et Cartographica Sinica, 2021, 50(2):215-225. DOI:10.11947/j.AGCS.2021.20220095.
[8] LI S Z. Markov random field modeling in image analysis[M]. London:Springer Science & Business Media, 2009.
[9] LI Yan, DAI Jicheng, TAN Junxiang, et al. Global fine registration of point cloud in LiDAR SLAM based on pose graph[J]. Journal of Geodesy and Geoinformation Science, 2020, 3(2):26-35.
[10] ANAND A, KOPPULA H S, JOACHIMS T, et al. Contextually guided semantic labeling and search for 3D point clouds[J]. International Journal of Robotics Research, 2011, 32(1):19-34.
[11] MUNOZ D, VANDAPEL N, HEBERT M. Onboard contextual classification of 3D point clouds with learned high-order Markov random fields[C]//Proceedings of 2009 IEEE International Conference on Robotics and Automation. Kobe, Japan:IEEE, 2009.
[12] LI Zhuqiang, ZHANG Liqiang, ZHONG Ruofei, et al. Classification of urban point clouds:a robust supervised approach with automatically generating training data[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2016, 10(3):1207-1220.
[13] WANG Yan, JI Rongrong, CHANG Shifu. Label propagation from imagenet to 3D point clouds[C]//Proceedings of 2013 IEEE conference on computer vision and pattern recognition. Portland, OR, USA:IEEE, 2013:3135-3142.
[14] FLOROS G, LEIBE B. Joint 2D-3D temporally consistent semantic segmentation of street scenes[C]//Proceedings of 2012 IEEE Conference on Computer Vision and Pattern Recognition. Providence, RI, USA:IEEE, 2012:2823-2830.
[15] FU Yifan, ZHU Xingquan, LI Bin. A survey on instance selection for active learning[J]. Knowledge and Information Systems, 2013, 35(2):249-283.
[16] WANG Yu, MENDEZ A E M, CARTWRIGHT M, et al. Active learning for efficient audio annotation and classification with a large amount of unlabeled data[C]//Proceedings of 2019 ICASSP International Conference on Acoustics, Speech and Signal Processing. Brighton, UK:IEEE, 2019:880-884.
[17] THOMAS H, QI C R, DESCHAUD J E, et al. Kpconv:flexible and deformable convolution for point clouds[C]//Proceedings of 2019 IEEE/CVF International Conference on Computer Vision. Seoul, South Korea:IEEE, 2019:6411-6420.
[18] YANG L, MACEACHREN A M, MITRA P, et al. Visually-enabled active deep learning for (geo) text and image classification:a review[J]. ISPRS International Journal of Geo-Information, 2018, 7(2):65.
[19] VEZHNEVETS A, BUHMANN J M, FERRARI V. Active learning for semantic segmentation with expected change[C]//Proceedings of 2012 IEEE conference on computer vision and pattern recognition. Providence, RI, USA:IEEE, 2012:3162-3169.
[20] LIN Y, VOSSELMAN G, CAO Y, et al. Active and incremental learning for semantic ALS point cloud segmentation[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2020, 169:73-92.
[21] 罗欢. 高分辨率激光扫描点云语义标注研究[D].厦门:厦门大学,2017. LUO Huan. Research on semantic labeling of high-resolution laser scanning point clouds[D]. Xiamen:Xiamen University,2017.
[22] LUO H, WANG C, Wen C, et al. Semantic labeling of mobile LiDAR point clouds via active learning and higher order MRF[J]. IEEE Transactions on Geoscience and Remote Sensing, 2018, 56(7):3631-3644.
[23] ADAMS A, GELFAND N, DOLSON J, et al. Gaussian kd-trees for fast high-dimensional filtering[M]. New York:ACM press. 2009:1-12.
[24] WEINMANN M, JUTZI B, HINZ S, et al. Semantic point cloud interpretation based on optimal neighborhoods, relevant features and efficient classifiers[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2015, 105:286-304.
[25] MALLET C, BRETAR F, ROUX M, et al. Relevance assessment of full-waveform LiDAR data for urban area classification[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2011, 66(6):S71-S84.
[26] TAO Wuyong, HUA Xianghong, WANG Ruisheng, et al. Quintuple local coordinate images for local shape description[J]. Photogrammetric Engineering & Remote Sensing, 2020, 86(2):121-132.
[27] TAO Wuyong, HUA Xianghong, YU Kegen, et al. A pipeline for 3D object recognition based on local shape description in cluttered scenes[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 59(1):801-816.
[28] WEINMANN M, JUTZI B, MALLET C. Feature relevance assessment for the semantic interpretation of 3D point cloud data[J]. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2013, 5(W2):1.
[29] GUYON I, WESTON J, BARNHILL S, et al. Gene selection for cancer classification using support vector machines[J]. Machine learning, 2002, 46(1):389-422.
[30] GR MPING U. Variable importance assessment in regression:linear regression versus random forest[J]. The American Statistician, 2009, 63(4):308-319.
[31] CARDOSO T N C, SILVA R M, CANTUO S, et al. Ranked batch-mode active learning[J]. Information Sciences, 2017, 379:313-337.
[32] LEWIS D D, GALE W A. A sequential algorithm for training text classifiers[C]//Proceedings of 1994 SIGIR. London, UK:Springer,1994:3-12.
[33] BRANDT J. Transform coding for fast approximate nearest neighbor search in high dimensions[C]//Proceedings of 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. San Francisco, CA, USA:IEEE, 2010:1815-1822.
[34] XU S, WANG R, WANG H, et al. An optimal hierarchical clustering approach to mobile LiDAR point clouds[J]. IEEE Transactions on Intelligent Transportation Systems, 2019, 21(7):2765-2776.
[35] HACKEL T, SAVINOV N, LADICKY L, et al. Semantic 3D.net:a new large-scale point cloud classification benchmark[J]. ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Sciences, 2017(IV-1/W1):91-98.
[36] PATRA S, BRUZZONE L. A batch-mode active learning technique based on multiple uncertainty for SVM classifier[J]. IEEE Geoscience and Remote Sensing Letters, 2011, 9(3):497-501.

点云场景语义标注的排序批处理模式主动学习法

Ranked batch-mode active learning method for semantic annotation of point cloud scene

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 6

编辑推荐

Metrics

本文评价

[1]	董卫华, 王圣凯, 王雪元, 杨天宇. 地图线状要素眼动识别的朴素贝叶斯方法[J]. 测绘学报, 2021, 50(6): 749-756.
[2]	赵生银, 安如, 朱美如. 联合像元-深度-对象特征的遥感图像城市变化检测[J]. 测绘学报, 2019, 48(11): 1452-1463.
[3]	熊艳, 高仁强, 徐战亚. 机载LiDAR点云数据降维与分类的随机森林方法[J]. 测绘学报, 2018, 47(4): 508-518.
[4]	田彦平, 陶超, 邹峥嵘, 杨钊霞, 何小飞. 主动学习与图的半监督相结合的高光谱影像分类[J]. 测绘学报, 2015, 44(8): 919-926.
[5]	左志权,张祖勋,张剑清. 一种基于多分辨率拓扑网络的高质量初始图斑对象生成方法[J]. 测绘学报, 2012, 41(1): 65-0.
[6]	虞欣. 贝叶斯网络在航空影像纹理分类中的应用研究[J]. 测绘学报, 2009, 38(4): 0-310.