基于点代理增强和逐层上采样的猪体点云补全方法

doi:10.16174/j.issn.1673-4645.2024.01.012

摘要/Abstract

摘要： 采用逆向工程技术进行猪体的三维重建并测算，是低成本无接触式猪体型体况评估的一大解决方案，在比较单视角和多视角采集方法的优缺点后，本文提出基于深度学习的点云补全方法，将猪体局部点云恢复成一个完整的点云以实现猪体三维重建。该猪体点云补全方法基于点代理增强和逐层上采样，首先通过特征提取结合位置嵌入生成点代理，使用点代理增强Transformer进一步提高点代理的特征表示能力，再基于点代理通过逐层上采样由粗到细逐步恢复最终的高分辨率、细粒度和分布均匀的完整点云。本文对实际生产环境中采集的猪体点云进行补全，所提方法与目前主流的点云补全方法进行对比试验，在多个指标的评定上，本文提出的方法都取得了较好性能，尤其是在猪体点云缺失严重补全难度较大的情况下效果更为突出。试验证明该方法对猪体主干部位的补全具备应用价值，能够用于实现基于局部点云的猪体三维点云重建。

关键词: 猪, 三维重建, 深度学习, 猪体点云补全, Transformer, 点云上采样

尹令, 罗泗港, 吴珍芳, 蔡更元, 沈卓婷, 李钦萍, 周润林. 基于点代理增强和逐层上采样的猪体点云补全方法[J]. 中国猪业, 2024, 19(1): 84-89.

参考文献

[1]司永胜, 安露露, 刘刚, 等. 基于Kinect相机的猪体理想姿态检测与体尺测量[J]. 农业机械学报, 2019, 50(1):58-65. Si YS, An LL, Liu G, et al.Ideal Posture Detection and Body Size Measurement of Pig Based on Kinect[J]. Transactions of the Chinese Society for Agricultural Machinery, 2019, 50(1):58-65. [2]滕光辉, 申志杰, 张建龙, 等. 基于Kinect传感器的无接触式母猪体况评分方法[J]. 农业工程学报, 2018, 34(13): 211-217. Teng GH, Shen ZJ, Zhang JL, et al. Non-contact sow body condition scoring method based on Kinect sensor[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2018, 34(13): 211-217. [3] Shi S, Yin L, Liang SH, et al. Research on 3D surface reconstruction and body size measurement of pigs based on multi-view RGB-D cameras[J]. Computers and Electronics in Agriculture, 2020, 175: 105543-105552. [4] 尹令, 蔡更元, 田绪红, 等. 多视角深度相机的猪体三维点云重构及体尺测量[J]. 农业工程学报, 2019, 35(23):201-208. Yin L, Cai GY, Tian XH, et al. Three dimensional point cloud reconstruction and body size measurement of pigs based on multi-view depth camera[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2019, 35(23): 201-208. [5]刘同海, 滕光辉, 张盛南, 等. 基于点云数据的猪体曲面三维重建与应用[J]. 农业机械学报, 2014, 45(6):291-295. Liu TH, Teng GH, Zhang SN, et al. Reconstruction and Application of 3D Pig Body Model Based on Point Cloud Data[J]. Transactions of the Chinese Society for Agricultural Machinery, 2014, 45(6):291-295. [6] Yuan W, Khot T, Held D, et al. Pcn: point completion network[C]//In: 2018 international conference on 3D vision (3DV). IEEE, 2018. [7] Tchapmi LP, Kosaraju V, Rezatofighi H, et al. Topnet: structural point cloud decoder[C]//In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019. [8] Xie H, Yao H, Zhou S, et al. Grnet: Gridding residual network for dense point cloud completion[C]//In: Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part IX. Cham: Springer International Publishing, 2020. [9] Xiang P, Wen X, Liu YS, et al. Snowflakenet: point cloud completion by snowflake point deconvolution with skip-transformer[C]//In: Proceedings of the IEEE/CVF international conference on computer vision, 2021. [10] Yu X, Rao Y, Wang Z, et al. Pointr: Diverse point cloud completion with geometry-aware transformers[C]//In: Proceedings of the IEEE/CVF international conference on computer vision, 2021. [11] Yu X, Rao Y, Wang Z, et al. AdaPoinTr: Diverse point cloud completion with adaptive geometry-aware transformers[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2023, 45(12): 14114-14130. [12] Zhou H, Cao Y, Chu W, et al. Seedformer: patch seeds based point cloud completion with upsample transformer[C]//In: Computer Vision–ECCV 2022: 17th European Conference, Tel Aviv, Israel, October 23–27, 2022, Proceedings, Part III. Cham: Springer Nature Switzerland, 2022. [13] 曾安, 彭杰威, 刘畅, 等. 基于多尺度几何感知Transformer的植物点云补全网络[J]. 农业工程学报, 2022, 38(4): 198-205. Zeng A, Peng JW, Liu C, et al. Plant point cloud completion network based on multi-scale geometry-aware point Transformer[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2022, 38(4): 198-205. [14] Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need[J]. arXiv, 2017. doi: 10.48550/arXiv.1706.03762. [15] Qi CR, Yi L, Su H, et al. PointNet++: deep hierarchical feature learning on point sets in a metric space[J]. 2017. doi: 10.48550/arXiv.1706.02413. [16] Zhao H, Jiang L, Jia J, et al. Point transformer[C]//In: Proceedings of the IEEE/CVF international conference on computer vision, 2021. [17] Wang Y, Sun Y, Liu Z, et al. Dynamic graph cnn for learning on point clouds[J]. Acm Transactions On Graphics (tog), 2019, 38(5):1-12. [18] Fan H, Su H, Guibas LJ. A point set generation network for 3d object reconstruction from a single image[C]//In: Proceedings of the IEEE conference on computer vision and pattern recognition, 2017. [19] Wu T, Pan L, Zhang JZ, et al. Balanced chamfer distance as a comprehensive metric for point cloud completion[J]. Advances in Neural Information Processing Systems, 2021, 34: 29088-29100.

相关文章 15

[1]	周业勋, 李凯, 钟发刚, 张宏福. 开发油莎豆新型饲料资源大有潜力[J]. 中国猪业, 2024, 19(1): 3-10.
[2]	李敬, 毕庆悦, 金英海, 蒋显仁, 李习龙. 25-羟基维生素D3的生物学特性及其在猪禽生产中的应用[J]. 中国猪业, 2024, 19(1): 11-18.
[3]	李雨菁，郝瑞荣. 葡萄籽原花青素对不同生长阶段猪生长性能的影响[J]. 中国猪业, 2024, 19(1): 19-24.
[4]	赵勤辉，许婷婷，刘家，汤海鸥. 复合营养性膏剂对新生仔猪生长性能的影响[J]. 中国猪业, 2024, 19(1): 25-31.
[5]	黄志洋，淡海锋，沈林園，朱砺，甘麦邻. 常见商业化猪精子计数板使用效果与测试分析[J]. 中国猪业, 2024, 19(1): 32-38.
[6]	陶璇，梁艳，杨雪梅，汪勇，杨跃奎，孔繁晶，王言，龚建军，朱康平，卫纪，杨少鹏，吕学斌，顾以韧，何志平. 中外猪种间生长发育相关的差异候选基因筛选[J]. 中国猪业, 2024, 19(1): 39-46.
[7]	邹兴启, 李芳韬, 刘业兵. 我国猪瘟流行现状及净化之路[J]. 中国猪业, 2024, 19(1): 47-52.
[8]	张秦川, 李云辉, 张满义, 薛嘉熹, 孙雪梅, 张自瑞, 乔亚萱, 刘伟楠, 蒋松, 肖非, 高宏伟, 孙延鸣, 盛金良, 张彦兵. 金银花源miR2911靶向PEDV基因区域分析[J]. 中国猪业, 2024, 19(1): 53-56.
[9]	白雪，莫玉鹏，李茂宁，郑浩东，陈思宇，王晓晔. 一例规模化猪场流行性腹泻病的诊断与防控风险分析[J]. 中国猪业, 2024, 19(1): 57-62.
[10]	吴雨清, 王文赞, 朱志平. 我国生猪粪尿养分资源测算及土地承载力分析[J]. 中国猪业, 2024, 19(1): 63-72.
[11]	冯兴尧, 王海峰, 朱君, 孙想, 邱阳, 李斌. 基于YOLOv5模型的仔猪社交识别方法研究[J]. 中国猪业, 2024, 19(1): 73-83.
[12]	张海峰，王林，陈南，张文涛，王祖力. 生猪“保险＋期货”模式中的成功案例[J]. 中国猪业, 2024, 19(1): 90-94.
[13]	鲍艳珍. 我国大规模生猪养殖全要素生产率的时空差异——基于Malmquist指数和超效率SBM模型[J]. 中国猪业, 2024, 19(1): 95-101.
[14]	谭莹. 价格下行期的生猪产业调研及预测分析[J]. 中国猪业, 2023, 18(6): 13-18.
[15]	张海峰, 陈南, 黄菊霞, 王林, 王祖力. 生猪期货上市对我国生猪产业的影响[J]. 中国猪业, 2023, 18(6): 19-22.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed