Detection of wheat seedling lines in the complex environment via deep learning

Haibo Lin, Yuandong Lu, Rongcheng Ding, Yufeng Xiu, Fazhan Yang

Abstract


Wheat seedling line detection is critical for precision agriculture and automatic guidance in early wheat field operation. Aiming at the complex wheat field environment, a method of detecting wheat seedling lines based on deep learning was proposed in this study. Firstly, a rotated bounding box was created to improve the YOLOv3 model to predict the approximate position of the wheat seedling line; Then, according to the rotated bounding region obtained by the model, the wheat seedling line was detected by fitting the extracted center points. Finally, a comprehensive evaluation method combining angle error and distance error was proposed to evaluate the accuracy of the extracted crop line. By testing images of wheat seedlings in different environments, the results showed that the mean angle error and distance error respectively reached 0.75° and 10.84 pixels while the mean running time was 63.83 ms for a 1920×1080 pixels image. And compared to the original model the improved algorithm model improved the mAP value by 13.2%. The angle error and the distance error of the improved algorithm model were reduced by 51.4% and 39.7%, respectively. The method proposed in this study can accurately detect the wheat seedling lines at different stages and it is also suitable for the environments with weeds, shadow, bright light, and dark light. At the same time, it has a certain adaptability to wheat seedling images with a yaw angle in the shooting process. The research results could provide a reference for the automatic guidance of early wheat field machinery.
Keywords: wheat seedling lines, automatic guidance, deep learning, rotated bounding box, evaluation method
DOI: 10.25165/j.ijabe.20241705.7834

Citation: Lin H B, Lu Y D, Ding R C, Xiu Y F, Yang F Z. Detection of wheat seedling lines in the complex environment via deep learning. Int J Agric & Biol Eng, 2024; 17(5): 255-265.

Keywords


wheat seedling lines, automatic guidance, deep learning, rotated bounding box, evaluation method

Full Text:

PDF

References


Mueller N D, Gerber J S, Johnston M, Ray D K, Ramankutty N, Foley J A. Correction: Corrigendum: Closing yield gaps through nutrient and water management. Nature, 2013; 494: 390.

Yu S L, Qi X H, Liu X Y, Dong Q Y, Xu Y M. Researches on the effect on yield increase by deep cultivation-root cutting in winter wheat. Scientia Agricultura Sinica, 1985; 4: 30–35. （in Chinese）

Li X H, Yang S X, Yu Z W, Yu S L. Effects of organic manure application on growth and senescence of root in winter wheat. Journal of Plant Nutrition and Fertilizers, 2005; 4: 467–472. （in Chinese）

Khaliq A, Shakeel M, Matloob A, Hussain S, Tanveer A, Murtaza G. Influence of tillage and weed control practices on growth and yield of wheat. Philippine Journal of Crop Science, 2013; 38（3）: 54–62.

Lv Z Q, Li R S, Dong Q Y, Yin K R, Yu S L. Mechanical root-cutting models for increasing the yield of winter wheat. Transactions of the CSAE, 2006; 4: 103–106. （in Chinese）

Xie B, Wu Z B, Mao E R. Development and prospect of key technologies on agricultural tractor. Transactions of the CSAM, 2018; 49: 1–17. （in Chinese）

Zhang M, Ji Y H, Li S C, Cao R Y, Xu H Z, Zhang Z Q. Research progress of agricultural machinery navigation technology. Transactions of the CSAM, 2020; 51（4）: 1–18. （in Chinese）

Aravind K R, Raja P, Pérez-Ruiz M. Task-based agricultural mobile robots in arable farming: A review. Spanish Journal of Agricultural Research, 2017; 15: 1–16.

Li J B, Chen B Q, Liu Y. Image detection method of navigation route of cotton plastic film mulch planter. Transactions of the CSAM, 2014; 45（1）: 40–45. （in Chinese）

Meng Q K, Qiu R C, He J, Zhang M, Ma X D, Liu G. Development of agricultural implement system based on machine vision and fuzzy control. Computers and Electronics in Agriculture, 2015; 112: 128–138.

Zhao T, Noboru N, Yang L L, Ishii K, Chen J. Fast edge detection method for wheat field based on visual recognition. Transactions of the CSAM, 2016; 47（11）: 32–37. （in Chinese）

Zhang X Y, Li X N, Zhang B H, Zhou J, Tian G Z, Xiong Y J, et al. Automated robust crop-row detection in maize fields based on position clustering algorithm and shortest path method. Computers and Electronics in Agriculture, 2018; 154: 165–175.

Gong J L, Wang X X, Zhang Y F, Lan Y B. Extraction method of corn rhizome navigation lines based on edge detection and area localization. Transactions of the CSAM, 2020; 51（10）: 26–33. （in Chinese）

Zhai Z Q, Zhu Z X, Du Y F, Song Z H, Mao E R. Multi-crop-row detection algorithm based on binocular vision. Biosystems Engineering, 2016; 150: 89–103.

Basso M, Pignaton de Freitas E. A UAV guidance system using crop row detection and line follower algorithms. Journal of Intelligent and Robotic Systems: Theory & Application, 2020; 97（3-4）: 605–621. doi: 10.1007/s10846-019-01006-0.

Guan Z H, Chen K Y, Ding Y C, Wu C Y, Liao Q X. Visual navigation path extraction method in rice harvesting. Transactions of the CSAM, 2020; 51（1）: 19–28. （in Chinese）

Miguel Guerrero J, Jaime Ruz J, Pajares G. Crop rows and weeds detection in maize fields applying a computer vision system based on geometry. Computers and Electronics in Agriculture, 2017; 142: 461–472.

Jiang G Q, Wang X J, Wang Z H, Liu H M. Wheat rows detection at the early growth stage based on Hough transform and vanishing point. Computers and Electronics in Agriculture, 2016; 123: 211–223.

Zhang Q, Chen S J M E, Li B. A visual navigation algorithm for paddy field weeding robot based on image understanding. Computers and Electronics in Agriculture, 2017; 143: 66–78.

Gong J L, Wang X X, Zhang Y F, Lan Y B, Mostafa K. Navigation line extraction based on root and stalk composite locating points. Computers and Electrical Engineering, 2021; 92: 107115.

Jiang G Q, Wang Z H, Liu H M. Automatic detection of crop rows based on multi-ROIs. Expert Systems with Applications, 2015; 42（5）: 2429–2441.

Yang Y, Zhang B L, Zha J Y, Wen X, Chen L Q, Zhang T, et al. Real-time extraction of navigation lines between corn rows. Transactions of the CSAE, 2020; 36（12）: 162–171. （in Chinese）

He J, Zang Y, Luo X W, Zhao R M, He J, Jiao J K. Visual detection of rice rows based on Bayesian decision theory and robust regression least squares method. Int J Agric & Biol Eng, 2021; 14（1）: 199–206.

Zhou Y, Yang Y, Zhang B L, Wen X, Yue X, Chen L Q. Autonomous detection of crop rows based on adaptive multi-ROI in maize fields. Int J Agric & Biol Eng, 2021; 14（4）: 217–225.

Lu Y D, Lin H B, Ding R C, Xiu Y F. A method for extracting identification lines of early wheat seedling based on machine vision. Journal of Agricultural Mechanization Research, 2023; 45（8）: 1–9. （in Chinese）

LeCun Y, Bengio Y, Hinton G. Deep learning. Nature, 2015; 521: 436–444.

Kamilaris A, Prenafeta-Boldú F X. Deep learning in agriculture: A survey. Computers and Electronics in Agriculture, 2018; 147: 70–90.

Koirala A, Walsha K B, Wang Z L, McCarthy C. Deep learning - Method overview and review of use for fruit detection and yield estimation. Computers and Electronics in Agriculture, 2019; 162: 219–234.

Liu F C, Yang Y, Zeng Y M, Liu Z Y. Bending diagnosis of rice seedling lines and guidance line extraction of automatic weeding equipment in paddy field. Mechanical Systems and Signal Processing, 2020; 142: 106791.

Pang Y, Shi Y Y, Gao S C, Jiang F, Veeranampalayam-Sivakumar A-N, Thompson L, et al. Improved crop row detection with deep neural network for early-season maize stand count in UAV imagery. Computers and Electronics in Agriculture, 2020; 178: 105766.

Khan S, Tufail M, Khan M T, Khan Z A, Anwar S. Deep learning-based identification system of weeds and crops in strawberry and pea fields for a precision agriculture sprayer. Precision Agriculture, 2021; 22: 1711–1727.

Liu G X, Nouaze J C, Mbouembe P L T, Kim J H. YOLO-tomato: A robust algorithm for tomato detection based on YOLOv3. Sensors, 2020; 20（7）: 2145.

Zhang Q, Wang J H, Li B. Extraction method for centerlines of rice seedings based on YOLOv3 target detection. Transactions of the CSAM, 2020; 51（8）: 34–43. （in Chinese）

Ma J Q, Shao W Y, Ye H, Wang L, Wang H, Zheng Y B, et al. Arbitrary-oriented scene text detection via rotation proposals. IEEE Transactions on Multimedia, 2018; 20（11）: 3111–3122.

Neves G, Ruiz M, Fontinele J, Oliveira L. Rotated object detection with forward-looking sonar in underwater applications. Expert Systems with Applications, 2020; 140: 112870.

Koirala A, Walsh K B, Wang Z L, Anderson N. Deep learning for mango （Mangifera indica） panicle stage classification. Agronomy, 2020; 10（1）: 143.

Yu Y, Zhang K L, Liu H, Yang L, Zhang D X. Real-time visual localization of the picking points for a ridge-planting strawberry harvesting robot. IEEE Access, 2020; 8: 116556–116568.

Li L H, Zhou Z Q, Wang B, Miao L J, Zong H. A novel CNN-based method for accurate ship detection in HR optical remote sensing images via rotated bounding box. IEEE Transactions on Geoscience and Remote Sensing, 2021; 59（1）: 686–699.

Huang J H, Zhang H Y, Wang L, Zhang Z L, Zhao C M. Improved YOLOv3 model for miniature camera detection. Optics & Laser Technology, 2021; 142: 107133.




Copyright (c) 2024 International Journal of Agricultural and Biological Engineering

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.

2023-2026 Copyright IJABE Editing and Publishing Office