相关实验视频

Updated: Jan 13, 2026

Visualizing Early Infection Sites of Rice Blast Disease Magnaporthe oryzae on Barley Hordeum vulgare Using a Basic Microscope and a Smartphone

Visualizing Early Infection Sites of Rice Blast Disease Magnaporthe oryzae on Barley Hordeum vulgare Using a Basic Microscope and a Smartphone

Published on: March 17, 2023

使用YOLOv5 Bi-HIC模型与实例和上下文信息进行藻叶斑,茶色斑点和茶色灰斑点疾病的增强检测

Quoc-Hung Phan¹, Bryan Setyawan¹, The-Phong Duong²

¹Department of Mechanical Engineering, National United University, Miaoli 360302, Taiwan.

Plants (Basel, Switzerland)

|October 29, 2025

概括

这项研究介绍了YOLOv5 Bi-HIC,这是一个新的AI模型,用于检测茶叶疾病,如藻类叶斑和. 这种自动化方法为茶叶农场提供了比手工检查更快,更准确的替代方案.

科学领域:

农业科学农业科学
计算机视觉计算机视觉
植物病理学植物病理学

背景情况:

茶叶疾病对全球茶叶生产构成重大威胁.
手动检查茶叶的疾病是劳动密集型和低效的大型农场.

研究的目的:

开发和评估一种新的卷积神经网络 (CNN) 模型,用于自动检测主要茶叶疾病.
通过结合实例和上下文信息来改进现有的对象检测模型,以提高性能.

主要方法:

开发了一种经过修改的YOLOv5物体检测模型,命名为YOLOv5 Bi-HIC.
该模型在1091张茶叶图像上接受了训练,这些图像显示了藻类叶斑,茶棕色斑,茶灰色斑.
实例和上下文信息被整合到YOLOv5架构中,以提高检测准确度.

主要成果:

YOLOv5 Bi-HIC 模型实现了高性能指标:精度 (0.977),回忆 (0.943),F1 评分 (0.968) 和平均精度 (mAP) (0.96).
特定疾病的检测可靠性得分很高:藻叶斑 (0.94),茶棕斑 (0.98),茶灰斑 (0.92).

结论:

YOLOv5 Bi-HIC模型展示了一种非常准确和高效的实时检测茶叶疾病的方法.
这种人工智能驱动的工具可以在茶叶种植园的及时管理和干预策略中显著帮助,提高作物产量和质量.

关键词:

这是YOLOv5的.卷积神经网络是一种卷积神经网络.茶叶疾病茶叶疾病

更多相关视频

Specific and Accurate Detection of the Citrus Greening Pathogen Candidatus liberibacter spp. Using Conventional PCR on Citrus Leaf Tissue Samples

Specific and Accurate Detection of the Citrus Greening Pathogen Candidatus liberibacter spp. Using Conventional PCR on Citrus Leaf Tissue Samples

Published on: June 29, 2018

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Published on: October 27, 2023

相关概念视频

Light Acquisition

Light Acquisition

In order to produce glucose, plants need to capture sufficient light energy. Many modern plants have evolved leaves specialized for light acquisition. Leaves can be only millimeters in width or tens of meters wide, depending on the environment. Due to competition for sunlight, evolution has driven the evolution of increasingly larger leaves and taller plants, to avoid shading by their neighbors with contaminant elaboration of root architecture and mechanisms to transport water and nutrients.