Tomato leaf disease detection based on improved CenterNet algorithm

doi:10.3785/j.issn.1008-973X.2025.11.016

Journal of ZheJiang University (Engineering Science)

2025, Vol. 59

Issue (11): 2370-2378 DOI: 10.3785/j.issn.1008-973X.2025.11.016

Tomato leaf disease detection based on improved CenterNet algorithm

Ya LI1(

),Chen JIANG1,Hairui WANG1,Guifu ZHU2,3,*(

),Can HU1

1. Faculty of Information Engineering and Automation, Kunming University of Science and Technology, Kunming 650504, China
2. Information Construction Center, Kunming University of Science and Technology, Kunming 650504, China
3. Kunming University ofScience and Technology - Dawn Information Industry Limited Company AI Joint Research Center, Kunming 650504, China

Download:

HTML

PDF(3346KB) HTML
Export: BibTeX | EndNote (RIS)

Abstract

A tomato leaf disease detection model based on the improved CenterNet algorithm was proposed in order to address the false detection and missed detection phenomena in traditional tomato leaf disease detection. A feature fusion module that integrated the attention mechanism was constructed in order to enhance the model's cross-scale feature fusion capability. The multi-branch convolutional module RFB was added to the backbone network in order to expand the receptive field and enhance the ability to extract target features. The pyramid convolution PyConv was introduced into the backbone network to enhance the extraction of multi-scale features by calculating receptive fields of different scales and reduce information loss. Pruning optimization strategies were designed in order to reduce the impact of introducing modules on the number of model parameters and computational load. The test results showed that the accuracy rate, recall rate, mAP₅₀ and mAP_50:95 of the improved model reached 96.3%, 80.2%, 91.4% and 78.7% respectively. The proposed model can effectively improve the accuracy of tomato leaf disease detection, and the model has good generalization.

Key words： tomato leaf disease CenterNet feature fusion pyramid convolution multi-branch convolution

Received: 15 November 2024 Published: 30 October 2025

CLC:	TP 391
	S 24

Fund: 国家自然科学基金资助项目（61863016）.

Corresponding Authors: Guifu ZHU E-mail: 59515091@qq.com;zhuguifu@kust.edu.cn

	Service
	E-mail this article
	Add to my bookshelf
	Add to citation manager
	E-mail Alert
	RSS
	Articles by authors
	Ya LI
	Chen JIANG
	Hairui WANG
	Guifu ZHU
	Can HU

Cite this article:

Ya LI,Chen JIANG,Hairui WANG,Guifu ZHU,Can HU. Tomato leaf disease detection based on improved CenterNet algorithm. Journal of ZheJiang University (Engineering Science), 2025, 59(11): 2370-2378.

URL:

https://www.zjujournals.com/eng/10.3785/j.issn.1008-973X.2025.11.016 OR https://www.zjujournals.com/eng/Y2025/V59/I11/2370

基于改进CenterNet算法的番茄叶片病害检测

为了解决在传统番茄叶片病害检测中出现的误检和漏检现象，提出基于改进CenterNet算法的番茄叶片病害检测模型. 构建融合注意力机制的特征融合模块，增强模型的跨尺度特征融合能力. 在骨干网络中加入多分支卷积模块RFB，扩大感受野，加强对目标特征的提取能力. 在骨干网络中引入金字塔卷积PyConv，通过计算不同尺度的感受野来强化多尺度特征的提取，减少信息损失. 设计剪枝优化策略，减少引入模块给模型参数量和计算量带来的影响. 试验结果显示，改进后模型的准确率、召回率、mAP₅₀和mAP_50:95达到96.3%、80.2%、91.4%和78.7%. 利用提出的模型，能够有效地提升番茄叶片病害检测的准确性，模型具有良好的泛化性.

关键词： 番茄叶片病害, CenterNet, 特征融合, 金字塔卷积, 多分支卷积

Fig.1 Tomato leaf disease

Fig.2 Structure diagram of CenterNet

Fig.3 Modified CenterNet network structure

Fig.4 Flow diagram of BAM process

Fig.5 Structure diagram of RFB

Fig.6 Structure diagram of PyConv

Tab.1 Ablation study on feature fusion module with attention mechanism

Tab.2 Ablation experiment on different combination of improvement point

Fig.7 Model loss curve of ablation experiment

Fig.8 Comparison of heatmap before and after improvement

Fig.9 Visual comparison of disease detection before and after improvement

Tab.3 Performance comparison experiment of different network model

Tab.4 Detection result of model on different dataset


[1]	王会征, 孙良晨, 李新龙, 等基于改进YOLOv7-tiny的番茄叶片病虫害检测方法[J]. 农业工程学报, 2024, 40 (10): 194- 202 WANG Huizheng, SUN Liangchen, LI Xinlong, et al Tomato leaf disease and pest detection method based on improved YOLOv7-tiny[J]. Transactions of the Chinese Society of Agricultural Engineering, 2024, 40 (10): 194- 202

[2]	刘佳明, 张欣, 陈孝玉龙, 等基于改进MobileNetV2的番茄病害识别研究[J]. 南京农业大学学报, 2025, 48 (3): 724- 733 LIU Jiaming, ZHANG Xin, CHEN Xiaoyulong, et al Research on tomato disease recognition based on improved MobileNetV2[J]. Journal of Nanjing Agricultural University, 2025, 48 (3): 724- 733

[3]	李健, 王晨, 马振宇, 等. MobileNet-CAL: 基于迁移学习和注意力机制的番茄病虫害分类方法[EB/OL]. [2025-08-22]. https://doi.org/10.13229/j.cnki.jdxbgxb.20230828.

[4]	嵇春梅, 周鑫志, 叶烨华. 自然场景下的轻量化番茄病害检测模型[EB/OL]. [2025-08-22]. https://link.cnki.net/urlid/32.1148.s.20241009.0833.002.

[5]	刘拥民, 刘翰林, 石婷婷, 等一种优化的Swin Transformer番茄叶片病害识别方法[J]. 中国农业大学学报, 2023, 28 (4): 80- 90 LIU Yongmin, LIU Hanlin, SH Tingting, et al An optimized Swin Transformer-based method for tomato leaf disease recognition[J]. Journal of China Agricultural University, 2023, 28 (4): 80- 90

[6]	王海瑞, 赵江河, 吴蕾, 等针对CenterNet缺点的安全帽检测算法改进[J]. 湖南大学学报: 自然科学版, 2023, 50 (8): 125- 133 WANG Hairui, ZHAO Jianghe, WU Lei, et al Improvement of a hard hat detection algorithm addressing the shortcomings of CenterNet[J]. Journal of Hunan University: Natural Sciences, 2023, 50 (8): 125- 133

[7]	BRAHIMI M, BOUKHALFA K, MOUSSAOUI A Deep learning for tomato diseases: classification and symptoms visualization[J]. Applied Artificial Intelligence, 2017, 31 (4): 299- 315

[8]	PARK J, WOO S, LEE J, et al A simple and light-weight attention module for convolutional neural networks[J]. International Journal of Computer Vision, 2020, 128 (4): 783- 798 doi: 10.1007/s11263-019-01283-0

[9]	LIU S, HUANG D, WANG Y. Receptive field block net for accurate and fast object detection [C]//Proceedings of the European Conference on Computer Vision. Munich: Springer, 2018: 385−400.

[10]	DUTA I, LIU L, ZHU F, et al. Pyramidal convolution: rethinking convolutional neural networks for visual recognition [EB/OL]. (2021-09-28). https://arxiv.org/abs/2006.11538.

[11]	HU J, SHEN L, SUN G. Squeeze-and-excitation networks [C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 7132−7141.

[12]	WOO S, PARK J, LEE J, et al. CBAM: convolutional block attention module [C]//European Conference on Computer Vision. Cham: Springer, 2018: 3−19.

[13]	ZHANG H, GOODFELLOW I, METAXAS D, et al. Self-attention with efficient local attention [EB/OL]. (2021-09-28). https://arxiv.org/abs/1907.09190.

[14]	DUAN K, BAI S, XIE L, et al. CenterNet: keypoint triplets for object detection [C]//IEEE/CVF International Conference on Computer Vision. Seoul: IEEE, 2019: 6568−6577.

[15]	WANG C, BOCHKOVSKIY A, LIAO H. YOLOv7: trainable bag-of-freebies sets new state-of-the-art for real-time object detectors [C]//IEEE/CVF Conference on Computer Vision and Pattern Recognition. Vancouver: IEEE, 2023: 7464−7475.

[16]	刘鹏, 张天翼, 冉鑫, 等. 基于PBM-YOLOv8的水稻病虫害检测[J]. 农业工程学报, 2024, 40(20): 147−156. LIU Peng, ZHANG Tianyi, RAN Xin, et al. Detection of rice diseases and pests based on PBM-YOLOv8 [J/OL]. Transactions of the Chinese Society of Agricultural Engineering, 2024, 40(20): 147−156.

[17]	BATTULA B, ASHOK B, SPICA G, et al Parkinson's disease detection using modified ResNext deep learning model from brain MRI images[J]. Soft Computing, 2023, 27 (16): 11905- 11914 doi: 10.1007/s00500-023-08535-9

[18]	ZHAO Y, LV W, XU S, et al. DETRs beat YOLOs on real-time object detection [EB/OL]. (2023-04-17). https://arxiv.org/abs/2304.08069.

[19]	宋耀莲, 王粲, 李大焱, 等基于改进YOLOv5s的无人机小目标检测算法[J]. 浙江大学学报: 工学版, 2024, 58 (12): 2417- 2426 SONG Yaolian, WANG Can, LI Dayan, et al An improved YOLOv5s-based drone small target detection algorithm[J]. Journal of Zhejiang University: Engineering Science, 2024, 58 (12): 2417- 2426

[1]	Wenxin CHENG,Guanghui YAN,Wenwen CHANG,Baijing WU,Yaning HUANG. Channel-weighted multimodal feature fusion for EEG-based fatigue driving detection[J]. Journal of ZheJiang University (Engineering Science), 2025, 59(9): 1775-1783.

[2]	Chaoqun DONG,Zhan WANG,Ping LIAO,Shuai XIE,Yujie RONG,Jingsong ZHOU. Lightweight YOLOv5s-OCG rail sleeper crack detection algorithm[J]. Journal of ZheJiang University (Engineering Science), 2025, 59(9): 1838-1845.

[3]	Zhuguo ZHOU,Yujun LU,Liye LV. Improved YOLOv5s-based algorithm for printed circuit board defect detection[J]. Journal of ZheJiang University (Engineering Science), 2025, 59(8): 1608-1616.

[4]	Jiarui FU,Zhaofei LI,Hao ZHOU,Wei HUANG. Camouflaged object detection based on Convnextv2 and texture-edge guidance[J]. Journal of ZheJiang University (Engineering Science), 2025, 59(8): 1718-1726.

[5]	Shenchong LI,Xinhua ZENG,Chuanqu LIN. Multi-task environment perception algorithm for autonomous driving based on axial attention[J]. Journal of ZheJiang University (Engineering Science), 2025, 59(4): 769-777.

[6]	Zhenli ZHANG,Xinkai HU,Fan LI,Zhicheng FENG,Zhichao CHEN. Semantic segmentation algorithm for multiscale remote sensing images based on CNN and Efficient Transformer[J]. Journal of ZheJiang University (Engineering Science), 2025, 59(4): 778-786.

[7]	Dengfeng LIU,Wenjing GUO,Shihai CHEN. Content-guided attention-based lane detection network[J]. Journal of ZheJiang University (Engineering Science), 2025, 59(3): 451-459.

[8]	Yongfu HE,Shiwei XIE,Jialu YU,Siyu CHEN. Detection method for spillage risk vehicle considering cross-level feature fusion[J]. Journal of ZheJiang University (Engineering Science), 2025, 59(2): 300-309.

[9]	Gang XIAO,Dapeng LU,Wenbo ZHENG,Zhenbo CHENG,Yuanming ZHANG. Multivariable time series data anomaly detection method based on spatiotemporal graph attention network[J]. Journal of ZheJiang University (Engineering Science), 2025, 59(10): 2134-2143.

[10]	Huan LIU,Yunhong LI,Leitao ZHANG,Yue GUO,Xueping SU,Yaolin ZHU,Lele HOU. Identification of apple leaf diseases based on MA-ConvNext network and stepwise relational knowledge distillation[J]. Journal of ZheJiang University (Engineering Science), 2024, 58(9): 1757-1767.

[11]	Yi LIU,Yidan CHEN,Lin GAO,Jiao HONG. Lightweight road extraction model based on multi-scale feature fusion[J]. Journal of ZheJiang University (Engineering Science), 2024, 58(5): 951-959.

[12]	Yin CAO,Junping QIN,Tong GAO,Qianli MA,Jiaqi REN. Generative adversarial network based two-stage generation of high-quality images from text[J]. Journal of ZheJiang University (Engineering Science), 2024, 58(4): 674-683.

[13]	Huijuan ZHANG,Kunpeng LI,Miaoxin JI,Zhenjiang LIU,Jianjuan LIU,Chi ZHANG. UAV detection algorithm based on spatial correlation enhancement[J]. Journal of ZheJiang University (Engineering Science), 2024, 58(3): 468-479.

[14]	Qingjie QIAN,Junhe YU,Hongfei ZHAN,Rui WANG,Jian HU. Dimension prediction method of injection molded parts based on multi-feature fusion of DL-BiGRU[J]. Journal of ZheJiang University (Engineering Science), 2024, 58(3): 646-654.

[15]	Yuebo MENG,Bo WANG,Guanghui LIU. Multi-scale context-guided feature elimination for ancient tower image classification[J]. Journal of ZheJiang University (Engineering Science), 2024, 58(12): 2489-2499.

Viewed

Full text

Abstract

Cited

Shared

Discussed