Design and Development of a Precision Spraying Control System for Orchards Based on Machine Vision Detection

Area of Science:

• Agricultural Engineering
• Computer Vision
• Control Systems

Background:

• Traditional pesticide application in orchards is inefficient and environmentally harmful due to subjective judgment.
• There is a growing need for precision spraying technologies to optimize resource use in orchard management.

Purpose of the Study:

• To develop and evaluate a machine vision-based precision spraying control system for orchards.
• To improve pesticide application efficiency and reduce environmental pollution in orchard production.

Main Methods:

• Developed a canopy leaf wall area calculation using a multi-iteration GrabCut image segmentation algorithm.
• Established a spray volume calculation model and proposed a fuzzy adaptive control algorithm with an extended state observer (ESO).
• Designed flow and pressure controllers and conducted laboratory and field performance tests on citrus trees.

Main Results:

• The system reduced spray volume by up to 59.73% compared to continuous spraying and 30.24% compared to PID control.
• Pesticide utilization efficiency increased by up to 61.42% compared to continuous spraying and 26.8% compared to PID control.
• Demonstrated significant reductions in spray volume and increases in efficiency compared to traditional and PID control methods.

Conclusions:

• The machine vision-based precision spraying system offers a novel technical approach for efficient and environmentally friendly orchard management.
• The developed system enhances agricultural production efficiency, fruit quality, and promotes environmental protection.
• The study highlights the significant application value of intelligent control systems in modern agriculture.