Weed Detection in Maize Fields by UAV Images Based on Crop Row Preprocessing and Improved YOLOv4

被引:44
作者
Pei, Haotian [1 ,2 ,3 ]
Sun, Youqiang [3 ]
Huang, He [3 ]
Zhang, Wei [1 ,2 ,3 ]
Sheng, Jiajia [3 ,4 ]
Zhang, Zhiying [5 ]
机构
[1] Anhui Univ, Inst Phys Sci, Hefei 230601, Peoples R China
[2] Anhui Univ, Inst Informat Technol, Hefei 230601, Peoples R China
[3] Chinese Acad Sci, Inst Intelligent Machines, Hefei Inst Phys Sci, Hefei 230031, Peoples R China
[4] Univ Sci & Technol China, Sci Isl Branch, Grad Sch, Hefei 230026, Peoples R China
[5] Jiangsu Univ, Inst Sci & Technol Informat, Zhenjiang 212013, Jiangsu, Peoples R China
来源
AGRICULTURE-BASEL | 2022年 / 12卷 / 07期
关键词
crop row mask; weed detection; samples balance; YOLOv4; Meta-ACON; attention module; Soft-NMS; UAV images;
D O I
10.3390/agriculture12070975
中图分类号
S3 [农学(农艺学)];
学科分类号
0901 ;
摘要
Effective maize and weed detection plays an important role in farmland management, which helps to improve yield and save herbicide resources. Due to their convenience and high resolution, Unmanned Aerial Vehicles (UAVs) are widely used in weed detection. However, there are some challenging problems in weed detection: (i) the cost of labeling is high, the image contains many plants, and annotation of the image is time-consuming and labor-intensive; (ii) the number of maize is much larger than the number of weed in the field, and this imbalance of samples leads to decreased recognition accuracy; and (iii) maize and weed have similar colors, textures, and shapes, which are difficult to identify when an UAV flies at a comparatively high altitude. To solve these problems, we propose a new weed detection framework in this paper. First, to balance the samples and reduce the cost of labeling, a lightweight model YOLOv4-Tiny was exploited to detect and mask the maize rows so that it was only necessary to label weeds on the masked image. Second, the improved YOLOv4 was used as a weed detection model. We introduced the Meta-ACON activation function, added the Convolutional Block Attention Module (CBAM), and replaced the Non-Maximum Suppression (NMS) with Soft Non-Maximum Suppression (Soft-NMS). Moreover, the distributions and counts of weeds were analyzed, which was useful for variable herbicide spraying. The results showed that the total number of labels for 1000 images decrease by half, from 33,572 to 17,126. The improved YOLOv4 had a mean average precision (mAP) of 86.89%.
引用
收藏
页数:18
相关论文
共 36 条
[1]   Remotely Piloted Aircraft (RPA) in Agriculture: A Pursuit of Sustainability [J].
Ahmad, Ali ;
Ordonez, Javier ;
Cartujo, Pedro ;
Martos, Vanesa .
AGRONOMY-BASEL, 2021, 11 (01)
[2]  
[Anonymous], 2022, MOBILENETS EFFICIENT
[3]  
Bochkovskiy A., 2020, PREPRINT
[4]   Soft-NMS - Improving Object Detection With One Line of Code [J].
Bodla, Navaneeth ;
Singh, Bharat ;
Chellappa, Rama ;
Davis, Larry S. .
2017 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), 2017, :5562-5570
[5]   UAVs for Vegetation Monitoring: Overview and Recent Scientific Contributions [J].
de Castro, Ana I. ;
Shi, Yeyin ;
Maja, Joe Mari ;
Pena, Jose M. .
REMOTE SENSING, 2021, 13 (11)
[6]  
Djaman K., 2022, Environmental Challenges, V6, DOI [10.1016/j.envc.2021.100417, DOI 10.1016/J.ENVC.2021.100417]
[7]   Deep Learning-Based Object Detection System for Identifying Weeds Using UAS Imagery [J].
Etienne, Aaron ;
Ahmad, Aanis ;
Aggarwal, Varun ;
Saraswat, Dharmendra .
REMOTE SENSING, 2021, 13 (24)
[8]   BOLD: Bio-Inspired Optimized Leader Election for Multiple Drones [J].
Ganesan, Rajesh ;
Raajini, X. Mercilin ;
Nayyar, Anand ;
Sanjeevikumar, Padmanaban ;
Hossain, Eklas ;
Ertas, Ahmet H. .
SENSORS, 2020, 20 (11)
[9]   Tree Recognition on the Plantation Using UAV Images with Ultrahigh Spatial Resolution in a Complex Environment [J].
Guo, Xuzhan ;
Liu, Qingwang ;
Sharma, Ram P. ;
Chen, Qiao ;
Ye, Qiaolin ;
Tang, Shouzheng ;
Fu, Liyong .
REMOTE SENSING, 2021, 13 (20)
[10]   GhostNet: More Features from Cheap Operations [J].
Han, Kai ;
Wang, Yunhe ;
Tian, Qi ;
Guo, Jianyuan ;
Xu, Chunjing ;
Xu, Chang .
2020 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2020, :1577-1586