Vision-Based Automatic Irrigation System for Greenhouse Tomato Using Leaf Keypoints

View at publisher Summarise with VIRAT
Authors: Riku Ohnuma, Makoto Koike, Hiroshi Mineno
Cited by: 0
References: 7

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Approximately 70% of the world’s freshwater is used for agriculture, making efficient water management in agriculture a critical challenge for achieving sustainability. In tomato cultivation, sensor-based automatic irrigation methods are widely adopted, however they require costly sensors and specialized knowledge for implementation and maintenance. Therefore, this study proposes a vision-based automatic irrigation system that determines irrigation timing based on leaf wilting quantified using keypoint detection technology (YOLO-Pose). The system consists of an HDR camera, environmental sensors (temperature, humidity, and illuminance), and a Raspberry Pi 4B, enabling low-cost implementation without specialized knowledge. In a tomato cultivation field trial, the proposed system achieved a 1.53 Brix increase in sugar content and a 12.6% reduction in water usage, confirming its effectiveness.

Fields of Study & Subjects

Smart Agriculture and AI Irrigation Practices and Water Management Leaf Properties and Growth Measurement

Publication details

Journal / source
International Conference on Consumer Electronics
Pages
1–4
Publication date
3 February 2026
Language
EN