Design and test of Kinect-based variable spraying control system for orchards

Area of Science:

• Agricultural Engineering
• Robotics
• Precision Agriculture

Background:

• Variable-rate spraying is crucial for efficient pesticide application.
• Existing methods for tree information detection in orchards are inefficient and demanding.
• Orchard plant protection requires advanced automated solutions.

Purpose of the Study:

• To design and evaluate a variable spraying control system for orchards using a Kinect sensor.
• To improve the efficiency and precision of pesticide application in citrus orchards.
• To reduce pesticide consumption while ensuring effective pest control.

Main Methods:

• A variable spraying control system was designed using a Kinect sensor for real-time canopy volume detection.
• Pulse width modulation controlled solenoid valves for variable pesticide application.
• Canopy volume calculation and nozzle flow models were developed and validated.
• Field experiments assessed spraying effectiveness, deposition, and pesticide savings.

Main Results:

• The system achieved a maximum relative error of 10.54% in canopy volume detection.
• The nozzle flow model demonstrated a correlation coefficient exceeding 0.95.
• Variable-rate spraying showed good consistency between theoretical and actual spray volumes.
• Pesticide use was reduced by up to 57.14% with comparable or improved deposition density.

Conclusions:

• The Kinect sensor is feasible for guiding variable spraying operations in orchards.
• The developed system offers a practical solution for precise and efficient plant protection.
• This technology provides a valuable reference for future automated plant protection systems.