CQN-YOLO: A Lightweight and High-Precision Model for Corn Appearance Quality Detection
Keywords:
Food quality assessment, automated agricultural product, deep learning, corn appearance quality, YOLOv10nAbstract
This study advances food quality assessment by utilizing a lightweight deep learning model for the rapid and high-precision detection of corn appearance defects, addressing critical needs in the automated characterization of agricultural products. The proposed CornQualityNet (CQN-YOLO) is based on enhanced YOLOv10n. Firstly, four augmentation methods are employed to enhance the dataset, addressing the issues of overfitting and data insufficiency during training, and thereby improving the model's robustness. Secondly, the Slimneck structure is designed to optimize the YOLOv10n neck network, which achieves a lightweight model, reduces computational complexity, and improves computational speed. Lastly, the Slide Loss is used to enhance the classification ability of the model, which accelerates the convergence speed of the model and reduces computing time. The experimental results demonstrate that the proposed CQN-YOLO model achieves an accuracy of 97.8%, a recall of 96.6%, and a Mean Average Precision (mAP) of 99.0%, which are 3.5%, 1.3%, and 1.0% higher than the YOLOv10n model, respectively, and the GFLOPs are reduced from 6.5 to 5.9. Compared to other models in the field, the CQN-YOLO model performs best, enhancing the model's generalization ability, detection efficiency, and robustness while maintaining high precision and achieving model lightweightness. The model can be effectively used in detection equipment with limited computational resources and provides technical support for the detection of the appearance quality of other agricultural products.
References
[1] Standardization Administration of China. Corn (GB/T 1234-2024). Technical report, Standards Press of China, Beijing, 2018.
[2] U.S. Grains Council. 2024/2025 corn harvest quality report. Technical report, U.S. Grains Council, 2024.
[3] Diwakar Agarwal, Sweta, and P. Bachan. Machine learning approach for the classification of wheat grains. Smart Agricultural Technology, 3:100136, 2023.
[4] Mukesh Dalal and Payal Mittal. A systematic review of deep learning-based object detection in agriculture: Methods, challenges, and future directions. Computers, Materials and Continua, 84(1):57–91, 2025.
[5] Ang Wu, Juanhua Zhu, Yuli Yang, Xinping Liu, Xiushan Wang, Ling Wang, Hao Zhang, and Jing Chen. Classification of corn kernels grades using image analysis and support vector machine. Advances in Mechanical Engineering, 10, 2018.
[6] MA Jia-jia, WANG Ke-qiang, and ZHENG Yi-xiong. Research on the appearance quality detection and classification of peanut seeds based on machine vision. Journal of Anhui Agricultural Sciences, 49(10):4, 2021.
[7] GUO Dongdong, CAI Xiang, QI Jiandong, et al. Low-cost grading device for mango appearance quality using accelerated CNN with FPGA. Transactions of the Chinese Society of Agricultural Engineering, 40(21):137–145, 2024.
[8] Zhou Chun-Xina, Huo Yi-Zhi, Du You-Hai, Jiang Min-Lan, Zeng Ling-Guo, Zhang Chang-Jiang, and Shi Xiao-Wei. Non-destructive detection of soybean appearance quality based on improved RetinaNet. Journal of the Chinese Cereals and Oils Association, 39(9):172–180, 2024.
[9] Haoyang Wang, Guoxiong Zhou, and Guiyun Chen. CEA-Net: A multi-modal model for corn disease classification with dynamic fusion and cross-layer connection mechanism. Pattern Recognition, 173:112788, 2026.
[10] Joseph Redmon, Santosh Kumar Divvala, Ross B. Girshick, and Ali Farhadi. You only look once: Unified, real-time object detection. In 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 779–788, 2016.
[11] Gu Hai-Yang, Jiang Xue-Song, Shen Fei, Yu Xin-Zhi, and Zhou Hong-Ping. Non-destructive testing model of rice appearance quality. Journal of the Chinese Cereals and Oils Association, 38(11):230–237, 2023.
[12] FAN Xiaofei, WANG Linbai, LIU Jingyan, ZHOU Yuhong, ZHANG Jun, and SUO Xuesong. Corn seed appearance quality estimation based on improved YOLO v4. Transactions of the Chinese Society for Agricultural Machinery, 53(7):226–233, 2022.
[13] Ke Sun, Mengdi Tang, Shu Li, and Siyuan Tong. Mildew detection in rice grains based on computer vision and the YOLO convolutional neural network. Food Science & Nutrition, 12(2):860–868, 2024.
[14] Tirthankar Mohanty, Priyabrata Pattanaik, Subhaprada Dash, Hara Prasada Tripathy, and William Holderbaum. Smart robotic system guided with YOLOv5 based machine learning framework for efficient herbicide usage in rice (Oryza sativa L.) under precision agriculture. Computers and Electronics in Agriculture, 231:110032, 2025.
[15] Siqi Niu, Xiaolin Xu, Ao Liang, Yuliang Yun, Fengqi Hao, Jinqiang Bai, Wenyi Sheng, and Dexin Ma. High accuracy identification of maize seed varieties based on a lightweight improved YOLOv8. Scientific Reports, 15:18541, 2025.
[16] Hao Sun, Rui Fu, and Dae-Ki Kang. LigTomDet: Knowledge distillation in a new lightweight tomato disease detection model in planting fields. Pattern Recognition, 172:112628, 2026.
[17] Lei Fan, Yiwen Ding, Dongdong Fan, Donglin Di, Maurice Pagnucco, and Yang Song. GrainSpace: A large-scale dataset for fine-grained and domain-adaptive recognition of cereal grains. In 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 21084–21093, 2022.
[18] Xinle Gao, Zhiyong Xiao, and Zhaohong Deng. High accuracy food image classification via vision transformer with data augmentation and feature augmentation. Journal of Food Engineering, 365:111833, 2024.
[19] Ao Wang, Hui Chen, Lihao Liu, Kai Chen, Zijia Lin, Jungong Han, and Guiguang Ding. YOLOv10: Real-time end-to-end object detection. In A. Globerson, L. Mackey, D. Belgrave, A. Fan, U. Paquet, J. Tomczak, and C. Zhang, editors, Advances in Neural Information Processing Systems, volume 37, pages 107984–108011. Curran Associates, Inc., 2024.
[20] Hulin Li, Jun Li, Hanbing Wei, Zheng Liu, Zhenfei Zhan, and Qiliang Ren. Slim-neck by GSConv: A lightweight-design for real-time detector architectures. Journal of Real-Time Image Processing, 21(3):62.1–62.13, 2024.
[21] Amalia Luque, Alejandro Carrasco, Alejandro Martín, and Ana de las Heras. The impact of class imbalance in classification performance metrics based on the binary confusion matrix. Pattern Recognition, 91:216–231, 2019.
[22] Ziping Yu, Hongbo Huang, Weijun Chen, Yongxin Su, Yahui Liu, and Xiuying Wang. YOLO-FaceV2: A scale and occlusion aware face detector. Pattern Recognition, 155:110714, 2024.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Chenjia Qian, Azhi Zhang (Author)
This work is licensed under a Creative Commons Attribution 4.0 International License.
