Scholarlink

In recent times, Smart Agriculture is a topic that has received a lot of consideration in Vietnam. Intelligent agriculture machinery, crop growth environment monitoring, water-saving irrigation, and monitoring information about animal and plant life are just a few of many well-known uses of Internet of Things (IoT) in agriculture. This paper proposes an agricultural IoT system that can monitor environmental parameters, including air temperature, humidity, light intensity and soil moisture. Two YOLOv8 (You Only Look Once v8) models were embedded in the system for analyzing images captured from OV2640 camera and detecting four health statuses of mustard greens: healthy, abnormal, pest and torn. To achieve high precision and reliable detection, two YOLOv8 models were applied to two data analysis processes: the segmentation process and the detection process. The obtained results were then compared with those using only the detection process proving the effectiveness and applicability of the proposed system.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Monitoring and Detecting the Health Status of Mustard Greens Using YOLOv8 Model

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Giang Le Hoang,
  • Thuan Phung Van,
  • Quan Hoang Manh,
  • Hiep Dao Trong,
  • Hoa Tran Quang,
  • Khanh Duong Quang

摘要

In recent times, Smart Agriculture is a topic that has received a lot of consideration in Vietnam. Intelligent agriculture machinery, crop growth environment monitoring, water-saving irrigation, and monitoring information about animal and plant life are just a few of many well-known uses of Internet of Things (IoT) in agriculture. This paper proposes an agricultural IoT system that can monitor environmental parameters, including air temperature, humidity, light intensity and soil moisture. Two YOLOv8 (You Only Look Once v8) models were embedded in the system for analyzing images captured from OV2640 camera and detecting four health statuses of mustard greens: healthy, abnormal, pest and torn. To achieve high precision and reliable detection, two YOLOv8 models were applied to two data analysis processes: the segmentation process and the detection process. The obtained results were then compared with those using only the detection process proving the effectiveness and applicability of the proposed system.