An Intelligent Water-Saving Irrigation System Based on Multi-Sensor Fusion and Visual Servoing Control

TL;DR

This work tackles excessive water use in precision agriculture by presenting a compact robotic irrigation system that fuses embedded vision, precise actuation, and terrain stabilization. It combines a lightweight YOLOv8n vision pipeline on a K210, a simplified eye-in-hand calibration for a 3-DoF arm, and an adaptive PID leveling mechanism coordinated by an STM32 controller to operate reliably on uneven terrain. Experimental results across greenhouse, hillside, and variable lighting conditions show 30–50% water savings and water-use efficiency above 92%, while maintaining detection accuracy >96% and positioning errors <6 mm. The approach offers a scalable, cost-effective solution for smallholders and urban gardens, enabling real-time perception-action-stabilization for precise irrigation.

Abstract

This paper introduces an intelligent water-saving irrigation system designed to address critical challenges in precision agriculture, such as inefficient water use and poor terrain adaptability. The system integrates advanced computer vision, robotic control, and real-time stabilization technologies via a multi-sensor fusion approach. A lightweight YOLO model, deployed on an embedded vision processor (K210), enables real-time plant container detection with over 96% accuracy under varying lighting conditions. A simplified hand-eye calibration algorithm-designed for 'handheld camera' robot arm configurations-ensures that the end effector can be precisely positioned, with a success rate exceeding 90%. The active leveling system, driven by the STM32F103ZET6 main control chip and JY901S inertial measurement data, can stabilize the irrigation platform on slopes up to 10 degrees, with a response time of 1.8 seconds. Experimental results across three simulated agricultural environments (standard greenhouse, hilly terrain, complex lighting) demonstrate a 30-50% reduction in water consumption compared to conventional flood irrigation, with water use efficiency exceeding 92% in all test cases.

Figures (5)

• Figure 1: Overall system architecture and technical workflow. The blue section depicts the mechanical structure, which provides mobility and structural support. The orange section encompasses three core technologies: (1) enhanced visual recognition, designed for real-time plant detection; (2) robotic arm control, enabling precise water delivery; (3) adaptive leveling system, dedicated to terrain compensation. Additionally, experimental validation assesses the system’s performance across three agricultural scenarios. Data fusion and coordinated operation of all modules are managed by the STM32 main controller.
• Figure 2: Hand-Eye Calibration Configuration and Parameter Annotation for the 3-DoF Robotic Arm System. The diagram illustrates the spatial relationship between the camera frame $(X_c, Y_c, Z_c)$ mounted on the end-effector and the robot base frame $(X_0, Y_0, Z_0)$. Key parameters include: joint angles $\theta_1$, $\theta_2$, $\theta_3$; link lengths $L_1$, $L_2$; and the fixed transformation between the end-effector frame $(X_e, Y_e, Z_e)$ and camera frame.
• Figure 3: Visual Recognition Accuracy Under Different Environments
• Figure 4: Inference time and False Positive Rate Comparison
• Figure 5: Irrigation Volume and Water Use Efficiency Analysis