基于机器视觉的果树冠层探测技术研究进展

doi:10.3969/j.issn.1000-1565.2026.05.002

摘要/Abstract

摘要： 果园对靶变量施药技术旨在解决传统施药方法导致的农药残留和环境污染等问题. 利用机器视觉技术实时获取果树冠层轮廓、几何尺寸及生物量等特征信息,是实现果园对靶变量施药的前提.本文综述了国内外果树冠层探测技术发展现状,介绍了RGB相机和RGB-D相机2种常用的采集果树冠层图像的视觉传感器,对比分析了传统图像处理、深度学习和三维重建等技术在获取果树冠层特征信息的效果.最后,提出了当前机器视觉技术在果树冠层探测中的不足,未来的研究应重点关注多传感器融合、新网络架构引入、公开数据集平台完善、无人机喷药装备的研发等方面.

关键词: 机器视觉, 果树冠层, RGB-D相机, 深度学习

中图分类号:

S491

孙艺萌, 谷趁趁, 翟长远, 高振, 郝建军. 基于机器视觉的果树冠层探测技术研究进展[J]. 河北大学学报(自然科学版), 2026, 46(5): 462-474.

SUN Yimeng, GU Chenchen, ZHAI Changyuan, GAO Zhen, HAO Jianjun. Research progress on machine vision-based detection technology for fruit tree canopy[J]. Journal of Hebei University(Natural Science Edition), 2026, 46(5): 462-474.

参考文献

[1] 吴中勇,李延荣,董中丹.我国水果市场发展现状及对策研究[J].中国果菜, 2023, 43(11): 79-83.DOI:10.19590/j.cnki.1008-1038.2023.11.016.
[2] Wang Y F, Jia W D, Ou M X, et al. A review of orchard canopy perception technologies for variable-rate spraying[J]. Sensors, 2025, 25(16): 4898-4928. DOI:10.3390/s25164898.
[3] Koc C, Duran H, Gerdan Koc D. Orchard sprayer design for precision pesticide application[J]. Erwerbs-Obstbau, 2023, 65(5): 1819-1828. DOI:10.1007/s10341-023-00876-x.
[4] Nauman M, Ghafoor Dr A, Khan M, et al. Real time variable rate sprayer: Its development and comparative analysis of spray coverage with conventional orchard sprayer[J]. Pak J Agric Sci 2022, 59: 993-1002.DOI:10.21162/PAKJAS/22.89.
[5] Tumbo S D, Salyani M, Whitney J D, et al. Investigation of laser and ultrasonic ranging sensors for measurements of citrus canopy volume[J]. Appl Eng Agric, 2002, 18(3): 367-372. DOI:10.13031/2013.8587.
[6] Nan Y L, Zhang H C, Zheng J Q, et al. Low-volume precision spray for plant pest control using profile variable rate spraying and ultrasonic detection[J]. Front Plant Sci, 2023, 13: 1042769. DOI:10.3389/fpls.2022.1042769.
[7] Salas B, Salcedo R, Garcia-Ruiz F, et al. Design, implementation and validation of a sensor-based precise airblast sprayer to improve pesticide applications in orchards[J]. Precis Agric, 2024, 25(2): 865-888. DOI:10.1007/s11119-023-10097-7.
[8] Kaleem A, Aqib M, Saleem S R, et al. Feasibility of ultrasonic sensors in development of real-time plant canopy measurement system[J]. Environ Sci Proc, 2022, 23(1): 22-26. DOI:10.3390/environsciproc2022023022.
[9] Guo N, Xu N, Kang J M, et al. A study on canopy volume measurement model for fruit tree application based on LiDAR point cloud[J]. Agriculture, 2025, 15(2): 130-152. DOI:10.3390/agriculture15020130.
[10] Jiang S K, Qi P, Han L, et al. Navigation system for orchard spraying robot based on 3D LiDAR SLAM with NDT_ICP point cloud registration[J]. Comput Electron Agric, 2024, 220: 108870. DOI:10.1016/j.compag.2024.108870.
[11] Baltazar A R, Dos Santos F N,DeSousa M L, et al. 2D LiDAR-based system for canopy sensing in smart spraying applications[J]. IEEE Access, 2023, 11: 43583-43591. DOI:10.1109/ACCESS.2023.3271973.
[12] Lu Z A, Qi L J, Zhang H, et al. Image segmentation of UAV fruit tree canopy in a natural illumination environment[J]. Agriculture, 2022, 12(7): 1039-1054. DOI:10.3390/agriculture12071039.
[13] Cong P C, Zhou J C, Li S D, et al. Citrus tree crown segmentation of orchard spraying robot based on RGB-D image and improved mask R-CNN[J]. Appl Sci, 2023, 13(1): 164-183. DOI:10.3390/app13010164.
[14] Xue X Y, Luo Q, Ji Y H, et al. Design and test of Kinect-based variable spraying control system for orchards[J]. Front Plant Sci, 2023, 14: 1297879. DOI:10.3389/fpls.2023.1297879.
[15] Chehreh B, Moutinho A, Viegas C. Latest trends on tree classification and segmentation using UAV data: areview of agroforestry applications[J]. Remote Sens, 2023, 15(9): 2263-2302. DOI:10.3390/rs15092263.
[16] 罗锡文,吴海华,陈学庚,等.我国农业装备创新发展成就与未来展望[J].中国工程科学, 2026,28(1): 218-230. DOI:10.15302/J-SSCAE-2026.01.002.
[17] 曹冰雪,李鸿飞,赵春江,等.智慧农业科技创新引领农业新质生产力发展路径[J].智慧农业(中英文), 2024, 6(4): 116-127.
[18] 肖珂,郝毅,高冠东.果园自动变距精准施药系统设计与试验[J].农业机械学报, 2022, 53(10): 137-145.DOI: 10.6041/j.issn.1000-1298.2022.10.014.
[19] Ou M X, Hu T H, Hu M S, et al. Experiment of canopy leaf area density estimation method based on ultrasonic echo signal[J]. Agriculture, 2022, 12(10): 1569-1582. DOI:10.3390/agriculture12101569.
[20] Wang M M, Dou H J, Sun H Y, et al. Calculation method of canopy dynamic meshing division volumes for precision pesticide application in orchards based on LiDAR[J]. Agronomy, 2023, 13(4): 1077-1091. DOI:10.3390/agronomy13041077.
[21] 肖珂,王文静,高冠东,等.复杂背景下多种树形果树树冠的检测与识别方法研究[J].河北农业大学学报, 2022, 45(4): 100-108. DOI:10.13320/j.cnki.jauh.2022.0067.
[22] Bing Q, Zhang R, Zhang L, et al. UAV-SfM photogrammetry for canopy characterization toward unmanned aerial spraying systems precision pesticide application in an orchard[J]. Drones, 2025, 9(2): 151-181. DOI:10.3390/drones9020151.
[23] 薛秀云.基于激光雷达的果树冠层构建与变量喷雾技术研究[D].广州:华南农业大学, 2021.
[24] 刘慧,姜建滨,沈跃,等.基于改进DeepLab V3+的果园场景多类别分割方法[J].农业机械学报, 2022, 53(11): 255-261. DOI:10.6041/j.issn.1000-1298.2022.11.025.
[25] Sapkota R, Ahmed D, Karkee M. Comparing YOLOv8 and Mask R-CNN for instance segmentation in complex orchard environments[J]. ArtifIntell Agric, 2024, 13: 84-99. DOI:10.1016/j.aiia.2024.07.001.
[26] Li H, Guo C L, Yang Z S, et al. Design of field real-time target spraying system based on improved YOLOv5[J]. Front Plant Sci, 2022, 13: 1072631. DOI:10.3389/fpls.2022.1072631.
[27] Maheswari P, Raja P, Apolo-Apolo O E, et al. Intelligent fruit yield estimation for orchards using deep learning based semantic segmentation techniques: a review[J]. Front Plant Sci, 2021, 12: 684328. DOI:10.3389/fpls.2021.684328.
[28] Zhao H T, Morgenroth J, Pearse G, et al. A systematic review of individual tree crown detection and delineation with convolutional neural networks(CNN)[J]. Curr For Rep, 2023, 9(3): 149-170. DOI:10.1007/s40725-023-00184-3.
[29] Sun D Z, Liu W K, Run M L,et al. Monocular vision for variable spray control system[J]. Int J Agric Biol Eng, 2022, 15(6): 206-215. DOI:10.25165/j.ijabe.20221506.7646.
[30] Jintasuttisak T, Edirisinghe E, Elbattay A. Deep neural network based date palm tree detection in drone imagery[J]. Comput Electron Agric, 2022, 192: 106560. DOI:10.1016/j.compag.2021.106560.
[31] 韩阳.基于复杂背景的叶面积测定方法研究[D].太谷:山西农业大学, 2021. DOI:10.7666/D03225602
[32] Ferro M V, Sørensen C G, Catania P. Comparison of different computer vision methods for vineyard canopy detection using UAV multispectral images[J]. Comput Electron Agric, 2024, 225: 109277. DOI:10.1016/j.compag.2024.109277.
[33] Lei L, Yang Q L, Yang L, et al. Deep learning implementation of image segmentation in agricultural applications: a comprehensive review[J]. Artif Intell Rev, 2024, 57(6): 149-207. DOI:10.1007/s10462-024-10775-6.
[34] P V V, Behera M D, Trivedi S, et al. Optimized horticulture species identification and crown segmentation using multi-modal vision transformers and uav imagery[J]. Social Science Research Network, 2025. DOI:10.2139/ssrn.5370592.
[35] Qi L J, Zhou J R, Wan J J, et al. Canopy recognition of cherry fruit tree based on SegNet network model[C] //SPIE11915, International Conference on Agricultural and Food Processing Technology(ICAFPT). 2021: 92-104.DOI: 10.1117/12.2605881.
[36] Anagnostis A, Tagarakis A C, Kateris D, et al. Orchard mapping with deep learning semantic segmentation[J]. Sensors, 2021, 21(11): 3813-3832.DOI:10.3390/S21113813.
[37] Li Z K, Deng X L, Lan Y B, et al. Fruit tree canopy segmentation from UAV orthophoto maps based on a lightweight improved U-Net[J]. Comput Electron Agric, 2024, 217: 108538. DOI:10.1016/j.compag.2023.108538.
[38] Wang J H, Xiong H Y. A fast and accurate segmentation model of lychee tree canopy from UAV remote sensing image based on big data and deep learning[J]. J Big Data, 2025, 12(1): 142-164. DOI:10.1186/s40537-025-01199-2.
[39] 何海清,周福阳,陈敏,等.耦合卷积神经网络与注意力机制的无人机摄影测量果树树冠分割方法[J].地球信息科学学报, 2023, 25(12): 2387-2401. DOI:10.12082/dqxxkx.2023.230370.
[40] Charisis C, Argyropoulos D. Deep learning-based instance segmentation architectures in agriculture: a review of the Scopes and challenges[J]. Smart Agric Technol, 2024, 8: 100448. DOI:10.1016/j.atech.2024.100448.
[41] Afsar M M, Bakhshi A D, Iqbal M S, et al. High-precision mango orchard mapping using a deep learning pipeline leveraging object detection and segmentation[J]. Remote Sens, 2024, 16(17): 3207-3226. DOI:10.3390/rs16173207.
[42] La Y J, Seo D, Kang J, et al. Deep learning-based segmentation of intertwined fruit trees for agricultural tasks[J]. Agriculture, 2023, 13(11): 2097-2115.DOI:10.3390/agriculture13112097.
[43] Pichhika H C, Subudhi P, Yerra R V P. Automated segmentation and counting of different varieties of mango trees in UAV-videos[C] //2025 6th International Conference on Recent Advances in Information Technology(RAIT). March 6-8, 2025. Dhanbad, India. IEEE, 2025: 1-7. DOI:10.1109/rait65068.2025.11089445.
[44] Zhao G N, Yang R Z, Jing X D, et al. Phenotyping of individual apple tree in modern orchard with novel smartphone-based heterogeneous binocular vision and YOLOv5s[J]. Comput Electron Agric, 2023, 209: 107814. DOI:10.1016/j.compag.2023.107814.
[45] 闫成功,徐丽明,袁全春,等.基于双目视觉的葡萄园变量喷雾控制系统设计与试验[J].农业工程学报, 2021, 37(11): 13-22. DOI:10.11975/j.issn.1002-6819.2021.11.002.
[46] Khatri N, Shinde G U. Computer vision and image processing for precision agriculture[J]. Cognitive behavior and human computer interaction based on machine learning algorithm, 2021: 241-263.DOI:10.1002/9781119792109.ch11.
[47] Zhang R R, Lian S K, Li L L, et al. Design and experiment of a binocular vision-based canopy volume extraction system for precision pesticide application by UAVs[J]. Comput Electron Agric, 2023, 213: 108197. DOI:10.1016/j.compag.2023.108197.
[48] Kim J, Seol J, Lee S, et al. An intelligent spraying system with deep learning-based semantic segmentation of fruit trees in orchards[C] //2020 IEEE International Conference on Robotics and Automation(ICRA). May 31 - August 31, 2020, Paris, France. IEEE, 2020: 3923-3929. DOI:10.1109/ICRA40945.2020.9197556.
[49] Román C, Jeon H, Zhu H P, et al. Stereo vision controlled variable rate sprayer for specialty crops: part III. effect of travel speeds on spray deposition and ground loss[J]. J ASABE, 2023, 66(6): 1469-1479. DOI:10.13031/ja.15699.
[50] Jafari Malekabadi A, Khojastehpour M. Optimization of stereo vision baseline and effects of canopy structure, pre-processing and imaging parameters for 3D reconstruction of trees[J]. Mach Vis Appl, 2022, 33(6): 87-101. DOI:10.1007/s00138-022-01333-7.
[51] Khan Z, Liu H, Shen Y, et al. Deep learning improved YOLOv8 algorithm: Real-time precise instance segmentation of crown region orchard canopies in natural environment[J]. Comput Electron Agric, 2024, 224: 109168. DOI:10.1016/j.compag.2024.109168.
[52] Wei P, Yan X J, Yan W T, et al. Precise extraction of targeted apple tree canopy with YOLO-Fi model for advanced UAV spraying plans[J]. Comput Electron Agric, 2024, 226: 109425. DOI:10.1016/j.compag.2024.109425.
[53] Jin T T, Kang S M, Kim N R, et al. Comparative analysis of CNN-based semantic segmentation for apple tree canopy size recognition in automated variable-rate spraying[J]. Agriculture, 2025, 15(7): 789-807. DOI:10.3390/agriculture15070789.
[54] Paudel A, Davidson J R, Grimm C, et al. Vision-based normalized canopy area estimation for variable nitrogen application in apple orchards[J]. Smart Agric Technol, 2023, 5: 100309. DOI:10.1016/j.atech.2023.100309.
[55] Kothawade G S, Chandel A K, Schrader M J, et al. High throughput canopy characterization of a commercial apple orchard using aerial RGB imagery[C] //2021 IEEE International Workshop on Metrology for Agriculture and Forestry(MetroAgriFor). November 3-5, 2021. Trento-Bolzano, Italy. IEEE, 2021: 177-181. DOI:10.1109/metroagrifor52389.2021.9628564.
[56] Sinha R, Quirós J J, Sankaran S, et al. High resolution aerial photogrammetry based 3D mapping of fruit crop canopies for precision inputs management[J]. Inf Process Agric, 2022, 9(1): 11-23. DOI:10.1016/j.inpa.2021.01.006.
[57] Vinci A, Brigante R, Traini C, et al. Geometrical characterization of hazelnut trees in an intensive orchard by an unmanned aerial vehicle(UAV)for precision agriculture applications[J]. Remote Sens, 2023, 15(2): 541-556. DOI:10.3390/rs15020541.
[58] Zhang W L, Peng X Y, Bai T T, et al. A UAV-based single-lens stereoscopic photography method for phenotyping the architecture traits of orchard trees[J]. Remote Sens, 2024, 16(9): 1570-1596. DOI:10.3390/rs16091570.
[59] Gil E, Campos J, Ortega P, et al. DOSAVIÑA: Tool to calculate the optimal volume rate and pesticide amount in vineyard spray applications based on a modified leaf wall area method[J]. Compu Electron Agric, 2019, 160: 117-130. DOI:10.1016/j.compag.2019.03.018.
[60] 肖珂,梁聪哲,夏伟光.基于改进YOLACT的果树叶墙区域实时检测方法[J].农业机械学报, 2023, 54(4): 276-284. DOI: 10.6041/j.issn.1000-1298.2023.04.028.
[61] Gao G D, Xiao K, Ma Y J. A leaf-wall-to-spray-device distance and leaf-wall-density-based automatic route-planning spray algorithm for vineyards[J]. Crop Prot, 2018, 111: 33-41. DOI:10.1016/j.cropro.2018.04.015.
[62] Wang S L, Wang W, Lei X H, et al. Canopy segmentation method for determining the spray deposition rate in orchards[J]. Agronomy, 2022, 12(5): 1195-1207. DOI:10.3390/agronomy12051195.
[63] Gao G D, Xiao K, Li J P. Precision spraying model based on kinect sensor for orchard applications[J]. Appl Eng Agric, 2018, 34(2): 291-298. DOI:10.13031/aea.12538.
[64] Xu S L, Zheng S Q, Rai R. Dense object detection based canopy characteristics encoding for precise spraying in peach orchards[J]. Comput Electron Agric, 2025, 232: 110097. DOI:10.1016/j.compag.2025.110097.
[65] Vélez S, Poblete-Echeverría C, Rubio J A, et al. Estimation of Leaf Area Index in vineyards by analysing projected shadows using UAV imagery[J]. OENO One, 2021, 55(4): 159-180. DOI:10.20870/oeno-one.2021.55.4.4639.
[66] 张鹏九,高越,刘中芳,等.采用果园喷雾施药机械施药时农药有效沉积率的计算方法[J].农药学学报,2020,22(2):277-284.DOI: 10.16801/j.issn.1008-7303.2020.0056.
[67] Ilniyaz O, Du Q Y, Shen H F, et al. Leaf area index estimation of pergola-trained vineyards in arid regions using classical and deep learning methods based on UAV-based RGB images[J]. Comput Electron Agric, 2023, 207: 107723. DOI:10.1016/j.compag.2023.107723.