面向嵌入式除草机器人的玉米田间杂草识别方法

doi:10.3778/j.issn.1002-8331.2211-0282

摘要/Abstract

摘要： 为了实现嵌入式除草机器人在玉米田间准确、快速的进行除草工作，提出了一种实时目标检测算法GBC-YOLOv5s。使用1×1卷积和深度可分离卷积的组合替代普通卷积，在不改变输出特征图大小的情况下减少主干网络产生的杂草冗余特征。设计了一种双向特征融合网络（S-BiFPN）增强特征提取能力，充分利用不同尺度的特征提高杂草检测速度，并将多通道结构与自注意力机制结合，通过对输入特征进行压缩与再加权，以加强对小目标的关注度。针对不同的环境构建MWeed数据集进行测试，结果表明，与现有Yolov5s、Faster RCNN等模型方法相比，GBC-YOLOv5s模型轻量化后的大小仅为3.3 MB，输入图像的检测耗时（GPU）达到15.6 ms，平均精度（mAP）达到96.3%，能够有效地提升目标检测速度和识别精度，为农业智能除草领域提供理论依据。

关键词: YOLOv5s, 目标识别, 模型压缩, 特征融合

Abstract: In order to ensure the accuracy and rapidity of the embedded weeding robot in the corn field, a real-time target detection algorithm based on GBC-Yolov5s is proposed. First, the combination of the 1×1 convolution and depth-separable convolution is used to replace the traditional convolution, which reduces the redundant features generated by the backbone network without changing the size of the output feature map. Secondly, a bidirectional feature fusion network (S-BiFPN) network is designed to enhance the ability of feature extraction, which can make full use of different scale features to improve the speed of weed detection and combine the multi-channel structure with the self-attention mechanism to enhance the attention of small targets by compressing and reweighting the input features. Finally, MWeed data sets are built for different environments to test the proposed algorithm. The results show that compared with the Yolov5s and Faster RCNN model algorithms, the size of the GBC-Yolov5s model after lightweight is only 3.3 MB, the detection time of the input image (GPU) reaches 15.6 ms, and the average accuracy (mAP) reaches 96.3%, which can effectively improve the target detection speed and recognition accuracy, and provide a theoretical basis for the field of intelligent agricultural weeding.

Key words: YOLOv5s, target identification, model compression, feature fusion

何全令, 杨静文, 梁晋欣, 傅雷扬, 滕杰, 李绍稳. 面向嵌入式除草机器人的玉米田间杂草识别方法[J]. 计算机工程与应用, 2024, 60(2): 304-313.

HE Quanling, YANG Jingwen, LIANG Jinxin, FU Leiyang, TENG Jie, LI Shaowen. Weed Identification Method in Corn Fields Applied to Embedded Weeding Robots[J]. Computer Engineering and Applications, 2024, 60(2): 304-313.

参考文献

[1] BOYDSTON R A, WILLIAMS M M. Sweet corn hybrid tolerance to weed competition under three weed management levels[J]. Renewable Agriculture and Food Systems, 2016, 31(4): 281-287.
[2] TURSUN N, DATTA A, SAKINMAZ M S, et al. The critical period for weed control in three corn (Zea mays L.) types[J]. Crop Protection, 2016, 90: 59-65.
[3] HUSSAIN Z, MARWAT K B, CARDINA J, et al. Xanthium strumarium L. impact on corn yield and yield components[J]. Turkish Journal of Agriculture and Forestry, 2014, 38(1): 39-46.
[4] 王广祥, 王曌, 王宏波, 等. 吉林省春玉米田杂草防控现状及除草剂减量控害应用技术研究[J]. 东北农业科学, 2020, 45(5): 47-49.
WANG G X, WANG Z, WANG H B, et al. Current weed control and application technology of herbicide reduction in spring maize field in Jilin Province[J]. Journal of Northeast Agricultural Sciences, 2020, 45(5): 47-49.
[5] PIASECKI C, OVEJERO R F L, PICOLI JUNIOR G J, et al. Control of Italian ryegrass and Alexandergrass in corn using different corn sowing date, pre- and post-emergent herbicides[J]. Bragantia, 2020, 79(3): 387-398.
[6] QUAN L, FENG H, LV Y, et al. Maize seedling detection under different growth stages and complex field environments based on an improved faster R-CNN[J]. Biosystems Engineering, 2019, 184(1): 1-23.
[7] 傅雷扬, 李绍稳, 张乐, 等. 田间除草机器人研究进展综述[J]. 机器人, 2021, 43(6): 751-768.
FU L Y, LI S W, ZHANG L, et al. Review on research progress of weeding Robot in the field[J]. Robot, 2021, 43(6): 751-768.
[8] KANAGASINGHAM S, EKPANYAPONG M, CHAIHAN R. Integrating machine vision-based row guidance with GPS and compass-based routing to achieve autonomous navigation for a rice field weeding robot[J]. Precision Agriculture, 2020, 21(4): 831-855.
[9] AKBARZADEH S, PAAP A, AHDEROM S, et al. Plant discrimination by support vector machine classifier based on spectral reflectance[J]. Computers and Electronics in Agriculture, 2018, 148: 250-258.
[10] 张春龙, 黄小龙, 耿长兴, 等. 智能锄草机器人系统设计与仿真[J]. 农业机械学报, 2011, 42(7): 196-199.
ZHANG C L, HUANG X L, GENG C X, et al. Design and simulation of intelligent weeding robot system[J]. Transactions of the Chinese Society for Agricultural Machinery, 2011, 42(7): 196-199.
[11] BAWDEN O, KULK J, RUSSELL R, et al. Robot for weed species plant-specific management[J]. Journal of Field Robotics, 2017, 34(6): 1179-1199.
[12] WANG A C, ZHANG W, WEI X H. A review on weed detection using ground-based machine vision and image processing techniques[J]. Computers and Electronics in Agriculture, 2019, 158: 226-240.
[13] WU Z N, CHEN Y J, ZHAO B, et al. Review of weed detection methods based on computer vision[J]. Sensors, 2021, 21(11): 3647.
[14] KHAN S, TUKFAIL M, KHAN M T, et al. Deep learning-based identification system of weeds and crops in strawberry and pea fields for a precision agriculture sprayer[J]. Precision Agriculture, 2021, 22(6): 1711-1727.
[15] CHECHLINSKI L, SIEMIATKOWSKA B, MAJEWSKI M. A system for weeds and crops identification-reaching over 10 FPS on raspberry pi with the usage of MobileNets, DenseNet and custom modifications[J]. Sensors, 2019, 19(17): 3787.
[16] 颜秉忠. 机器视觉技术在玉米苗期杂草识别中的应用[J]. 农机化研究, 2018, 40(3): 212-216.
YAN B Z. Application of machine vision technology in weed identification at seedling stage of maize[J]. Journal of Agricultural Mechanization Research, 2018, 40(3): 212-216.
[17] WANG X S, ZHANG H H, CHEN Y. Research on maize canopy center recognition based on nonsignificant color difference segmentation[J]. Plos One, 2018, 13(9): 0202366.
[18] 姜红花, 张传银, 张昭, 等. 基于Mask R-CNN的玉米田间杂草检测方法[J]. 农业机械学报, 2020, 51(6): 220-228.
JIANG H H, ZHANG C Y, ZHANG Z, et al. A method of weed detection in maize field based on Mask R-CNN[J]. Transactions of the Chinese Society for Agricultural Machinery, 2020, 51(6): 220-228.
[19] ETIENNE A, AHMAD A, AGGARWAL V, et al. Deep learning-based object detection system for identifying weeds using UAS imagery[J]. Remote Sensing, 2021, 13(24): 5182.
[20] CHECHLINSKI L, SIEMIATKOWSKA B, MAJEWSKI M. A system for weeds and crops identification-reaching over 10 FPS on raspberry pi with the usage of MobileNets, DenseNet and custom modifications[J]. Sensors, 2019: 19173787.
[21] WONG A, FAMUORI M, SHAFIEE M J, et al. YOLO Nano: a highly compact you only look once convolutional neural network for object detection[J]. arXiv:1910.01271, 2019.
[22] 张啸. 基于嵌入式平台的杂草识别与空间定位方法研究及实现[D]. 哈尔滨: 哈尔滨工业大学, 2021.
ZHANG X. Research and implementation of weed identification and spatial localization method based on embedded platform[D]. Harbin: Harbin Institute of Technology, 2021.
[23] HAN K, WANG Y, XU C, et al. GhostNets on heterogeneous devices via cheap opertions[J]. International Journal of Computer Vision, 2022, 130(4): 1050-1069.
[24] DU F J, JIAO S J. Improvement of lightweight convolutional neural network model based on YOLO algorithm and its research in pavement defect detection[J]. Sensors, 2022, 22(9): 3537.
[25] ENVELOPE F J A P, JAS B, ANM C, et al. Detection of mold on the food surface using YOLOv5[J]. Current Research in Food Science, 2021, 4: 724-728.
[26] 卓天天, 桑庆兵. 注意力机制与复合卷积在手写识别中的应用[J]. 计算机科学与探索, 2022, 16(4): 888-897.
ZHUO T T, SANG Q B. Application of attention mechanism and composite convolution in handwriting recognition [J]. Journal of Frontiers of Computer Science and Technology, 2022, 16(4): 888-897.
[27] WANG C Y, LIAO H Y M, WU Y H, et al. CSPNet: a new backbone that can enhance learning capability of CNN[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2020: 390-391.