Weed Species Classification in Wheat Fields Using CNN-Based Image Analysis
Keywords:
Convolutional Neural Networks (CNN), Crop management, Data augmentation, Deep learning, DenseNet121, EfficientNetB0, Feature-level fusion, Hybrid model, Image-based plant identification, Multi-class classification, Precision agriculture, Smart farming, Transfer learning, Weed detectionAbstract
Among the most enduring and damaging biological hazards in wheat farming are weeds, which compete with crops for resources like sunlight, water, nutrients, and space. The productivity and sustainability of agriculture can be significantly impacted by its unchecked development, which can cut crop production by up to 80%. Conventional weed management techniques, such hand weeding and sprinkling herbicide, are time-consuming, damaging to the environment, and ineffective financially. In order to overcome this difficulty, we provide a deep learning-based method for automatically classifying weed species through picture analysis. The hybrid Convolutional Neural Network (CNN) model presented in this study combines the advantages of two pretrained architectures, DenseNet121 and EfficientNetB0. The model uses a bespoke classification head made for multi-class prediction after extracting and concatenating deep features from both networks, hence utilizing feature-level fusion. A publicly available Wheat-Weed picture dataset with eight classes six weed species and two wheat varieties is used to train and verify the algorithm. A ReduceLROnPlateau scheduler is utilized to adaptively modify learning rates during training, and data augmentation approaches are utilized to improve generalization. According to experimental data, the hybrid model performs noticeably better in terms of accuracy and resilience than the solo EfficientNetB0 and DenseNet121 models. With a test accuracy of 92.58% and a test loss of 0.2596, the suggested method has great potential for species-specific, real-time weed identification in precision agriculture. This method offers a scalable, effective, and environmentally friendly answer to the problems associated with contemporary weed control.
References
R. Punithavathi, C. Rani, K. R. Sughashini, C. Kurangi, and S. P. Balamurugan, “Computer Vision and Deep Learning-enabled Weed Detection Model for Precision Agriculture,” Computer Systems Science and Engineering 2023, vol. 44, no. 3, pp. 2759–2774, Jan. 2023, doi: https://doi.org/10.32604/csse.2023.027647.
S. N, M. Sundaram, and R. Ranjan, “Weedspedia: Deep Learning-Based Approach for Weed Detection using R-CNN, YoloV3 and Centernet,”: 2023 International Conference on Quantum Technologies, Communications, Computing, Hardware and Embedded Systems Security (iQ-CCHESS), pp. 1–5, Sep. 2023, doi: https://doi.org/10.1109/iq-cchess56596.2023.10391389.
A. Rabbani and M. Hussain, “YOLOv1 to YOLOv10: A comprehensive review of YOLO variants and their application in the agricultural domain,” Arxiv (Cornell University), Jun. 2024, doi: https://doi.org/10.48550/arxiv.2406.10139.
M. H. M. Roslim, “Using Remote Sensing and an Unmanned Aerial System for Weed Management in Agricultural Crops: A Review,” Agronomy, vol. 11, no. 9, p. 1809, Sep. 2021, doi: https://doi.org/10.3390/agronomy11091809.
W. Tian, W. Li, H. Song, and H. Yue, “Analysis on the difference of regional high-quality development in Beijing-Tianjin-Hebei city cluster,” Procedia Computer Science, vol. 199, pp. 1184–1191, Jan. 2022, doi: https://doi.org/10.1016/j.procs.2022.01.150.
D. Su, H. Kong, Y. Qiao, and S. Sukkarieh, “Data augmentation for deep learning based semantic segmentation and crop-weed classification in agricultural robotics,” Computers and Electronics in Agriculture, vol. 190, p. 106418, Nov. 2021, doi: https://doi.org/10.1016/j.compag.2021.106418.
M. Guzel, B. Turan, I. Kadioglu, A. Basturk, and B. Sin, “Deep Learning for Image-Based Detection of Weeds from Emergence to Maturity in Wheat Fields,” Smart Agricultural Technology, vol. 9, pp. 100552–100552, Sep. 2024, doi: https://doi.org/10.1016/j.atech.2024.100552.
S. Li and H. Lideskog, “Forest Terrain Object Detection Based on RGB and Depth Information,” DIVA, 2021. https://www.diva-portal.org/smash/record.jsf?pid=diva2%3A1544566&dswid=8714
N. Hussain, “Application of deep learning to detect Lamb’s quarters (Chenopodium album L.) in potato fields of Atlantic Canada,” Computers and Electronics in Agriculture, vol. 182, p. 106040, Mar. 2021, doi: https://doi.org/10.1016/j.compag.2021.106040.
H. Mennan, K. Jabran, B. H. Zandstra, and F. Pala, “Non-Chemical Weed Management in Vegetables by Using Cover Crops: A Review,” Agronomy, vol. 10, no. 2, p. 257, Feb. 2020, doi: https://doi.org/10.3390/agronomy10020257.
A. N. Veeranampalayam Sivakumar et al., “Comparison of Object Detection and Patch-Based Classification Deep Learning Models on Mid- to Late-Season Weed Detection in UAV Imagery,” Remote Sensing, vol. 12, no. 13, p. 2136, Jul. 2020, doi: https://doi.org/10.3390/rs12132136.
K. Wang, X. Hu, H. Zheng, “Weed detection and recognition in complex wheat fields based on an improved YOLOv7,” Frontiers in Plant Science, vol. 15, Jun. 2024, doi: https://doi.org/10.3389/fpls.2024.1372237
D. Chen, Y. Lu, Z. Li, and S. Young, “Performance evaluation of deep transfer learning on multi-class identification of common weed species in cotton production systems,” Computers and Electronics in Agriculture, vol. 198, p. 107091, Jul. 2022, doi: https://doi.org/10.1016/j.compag.2022.107091.
X. Zhang, “Weed Identification in Soybean Seedling Stage Based on Optimized Faster R-CNN Algorithm,” Agriculture, vol. 13, no. 1, p. 175, Jan. 2023, doi: https://doi.org/10.3390/agriculture13010175.
R. Reedha, E. Dericquebourg, R. Canals, and A. Hafiane, “Transformer Neural Network for Weed and Crop Classification of High-Resolution UAV Images,” Remote Sensing, vol. 14, no. 3, p. 592, Jan. 2022, doi: https://doi.org/10.3390/rs14030592.
P. Lottes, J. Behley, A. Milioto, and C. Stachniss, “Fully Convolutional Networks with Sequential Information for Robust Crop and Weed Detection in Precision Farming,” IEEE Robotics and Automation Letters, vol. 3, no. 4, pp. 2870–2877, Oct. 2018, doi: https://doi.org/10.1109/lra.2018.2846289.
S. Hameed and I. Amin, “(PDF) Detection of Weed and Wheat Using Image Processing,” www.researchgate.net, 2018. https://www.researchgate.net/publication/330796498_Detection_of_Weed_and_Wheat_Using_Image_Processing
K. Xu, Y. Zhu, W. Cao, X. Jiang, “Multi-Modal Deep Learning for Weeds Detection in Wheat Field Based on RGB-D Images,” Frontiers in Plant Science, vol. 12, Nov. 2021, doi: https://doi.org/10.3389/fpls.2021.732968.
L. Elstone, K. Y. How, S. Brodie, M. Z. Ghazali, W. P. Heath, and B. Grieve, “High Speed Crop and Weed Identification in Lettuce Fields for Precision Weeding,” Sensors, vol. 20, no. 2, p. 455, Jan. 2020, doi: https://doi.org/10.3390/s20020455.
S. L. Madsen, S. K. Mathiassen, M. Dyrmann, M. S. Laursen, and L.-C. Paz, “Open Plant Phenotype Database of Common Weeds in Denmark,” Remote Sensing, vol. 12, no. 8, pp. 1246–1246, Apr. 2020, doi: https://doi.org/10.3390/rs12081246.
K. Sudars, J. Jasko, I. Namatevs, L. Ozola, and N. Badaukis, “Dataset of annotated food crops and weed images for robotic computer vision control,” Data in Brief, vol. 31, p. 105833, Aug. 2020, doi: https://doi.org/10.1016/j.dib.2020.105833
J. Champ, A. Mora‐Fallas, H. Goëau, E. Mata‐Montero, P. Bonnet, and A. Joly, “Instance segmentation for the fine detection of crop and weed plants by precision agricultural robots,” Applications in Plant Sciences, vol. 8, no. 7, Jul. 2020, doi: https://doi.org/10.1002/aps3.11373.
A. S. M. Mahmudul Hasan, F. Sohel, and D. Diepeveen, “A survey of deep learning techniques for weed detection from images,” Computers and Electronics in Agriculture, vol. 184, p. 106067, May 2021, doi: https://doi.org/10.1016/j.compag.2021.106067.
J. Justina Michael and M. Thenmozhi, “Evaluation of Deep Learning CNN Models with 24 Metrics Using Soybean Crop and Broad-Leaf Weed Classification,” Lecture Notes in Networks and Systems, vol. 184, pp. 71–87, 2023, doi: https://doi.org/10.1007/978-981-99-5166-6_6.
N. Rai, “Applications of deep learning in precision weed management: A review,” Computers and Electronics in Agriculture, vol. 206, p. 107698, Mar. 2023, doi: https://doi.org/10.1016/j.compag.2023.107698.
A. Subeesh, S. Bhole, K. Singh, N. S. Chandel, “Deep convolutional neural network models for weed detection in polyhouse grown bell peppers,” Artificial Intelligence in Agriculture, vol. 6, pp. 47–54, 2022, doi: https://doi.org/10.1016/j.aiia.2022.01.002.
X. Jin, T. Liu, Y. Chen, and J. Yu, “Deep Learning-Based Weed Detection in Turf: A Review,” Agronomy, vol. 12, no. 12, p. 3051, Dec. 2022, doi: https://doi.org/10.3390/agronomy12123051.
F. Garibaldi-Márquez, G. Flores, and L. M. Valentín-Coronado, “Leveraging deep semantic segmentation for assisted weed detection,” Journal of Agricultural Engineering, vol. 56, no. 2, Feb. 2025, doi: https://doi.org/10.4081/jae.2025.1741.
C. Yuan, A. Tony, R. Yin, K. Wang, and W. Zhang, “Tactile and Thermal Sensors Built from Carbon–Polymer Nanocomposites A Critical Review,” Sensors, vol. 21, no. 4, p. 1234, Feb. 2021, doi: https://doi.org/10.3390/s21041234.
