Topic

Technologies and technical equipment for agriculture and food industry

Volume

Volume 78 / No. 1 / 2026

Pages : 1442-1455

Metrics

Volume viewed 0 times

Volume downloaded 0 times

LIGHTWEIGHT DETECTION OF VISIBLE CUCUMBER DOWNY MILDEW AND POWDERY MILDEW LESIONS UNDER GREENHOUSE CONDITIONS USING AN IMPROVED YOLOV11N

温室条件下基于改进 YOLOV11N 的黄瓜霜霉病与白粉病可见病斑轻量化检测

DOI : https://doi.org/10.35633/inmateh-78-112

Authors

Qinyou SUN

1. School of Information Science and Engineering, Shandong Agricultural University, Tai’an 271018, China 2. Key Laboratory of Huanghuaihai Smart Agricultural Technology, Ministry of Agriculture and Rural Affairs, Tai’an 271018, China 3. Agricultural Big Data Research Center, Shandong Agricultural University, Tai’an 271018, China

Xingyu GAO

1. School of Information Science and Engineering, Shandong Agricultural University, Tai’an 271018, China 2. Key Laboratory of Huanghuaihai Smart Agricultural Technology, Ministry of Agriculture and Rural Affairs, Tai’an 271018, China 3. Agricultural Big Data Research Center, Shandong Agricultural University, Tai’an 271018, China

Fengyu LI

1. School of Information Science and Engineering, Shandong Agricultural University, Tai’an 271018, China 2. Key Laboratory of Huanghuaihai Smart Agricultural Technology, Ministry of Agriculture and Rural Affairs, Tai’an 271018, China 3. Agricultural Big Data Research Center, Shandong Agricultural University, Tai’an 271018, China

Xianyong MENG

1. School of Information Science and Engineering, Shandong Agricultural University, Tai’an 271018, China 2. Key Laboratory of Huanghuaihai Smart Agricultural Technology, Ministry of Agriculture and Rural Affairs, Tai’an 271018, China 3. Agricultural Big Data Research Center, Shandong Agricultural University, Tai’an 271018, China

(*) Jun YAN

1. School of Information Science and Engineering, Shandong Agricultural University, Tai’an 271018, China 2. Key Laboratory of Huanghuaihai Smart Agricultural Technology, Ministry of Agriculture and Rural Affairs, Tai’an 271018, China 3. Agricultural Big Data Research Center, Shandong Agricultural University, Tai’an 271018, China

(*) Pingzeng LIU

1. School of Information Science and Engineering, Shandong Agricultural University, Tai’an 271018, China 2. Key Laboratory of Huanghuaihai Smart Agricultural Technology, Ministry of Agriculture and Rural Affairs, Tai’an 271018, China 3. Agricultural Big Data Research Center, Shandong Agricultural University, Tai’an 271018, China

(*) Corresponding authors:

yanj2016@sdau.edu.cn |

Jun YAN

pzliu@sdau.edu.cn |

Pingzeng LIU

Abstract

Downy mildew and powdery mildew pose significant threats to greenhouse cucumber production; however, accurate lesion detection on edge devices remains challenging due to the small size, high density, and frequent occlusion of lesions. This study proposes HSLG-YOLO, a lightweight detector based on YOLOv11n, which integrates CAA-HGNet, MPE-FPN, and TPL-Head for greenhouse cucumber lesion detection and edge deployment. Experimental results on a self-collected dataset show that the model achieves a precision of 93.58%, recall of 91.88%, mAP@0.5 of 96.49%, and mAP@0.5:0.95 of 67.89%. The model size is reduced from 5.3 MB to 4.6 MB, and it achieves 9.777 FPS on the Jetson Orin Nano platform, enabling near-real-time greenhouse disease monitoring.

Abstract in Chinese

霜霉病和白粉病威胁温室黄瓜生产，但由于病斑小、密集且易受遮挡影响，边端设备上的精准检测仍较困难。本文提出一种基于 YOLOv11n 的轻量化检测模型 HSLG-YOLO，融合 CAA-HGNet、MPE-FPN 和 TPL-Head，用于温室黄瓜病斑检测与边端部署。在自建数据集上，该模型的 Precision、Recall、mAP@0.5 和 mAP@0.5:0.95 分别达到 93.58%、91.88%、96.49% 和 67.89%。模型大小由 5.3 MB 降至 4.6 MB，并在 Jetson Orin Nano 上达到 9.777 FPS，可支持近实时温室病害巡检。

Indexed in

Road