IoT-based Orchard Monitoring System and Method

IoT Monitoring System for Fruit Maturity Assessment Using HSV Binarization Image Analysis

This technology monitors crop environments and growth status using cameras and temperature/humidity sensors mounted on a mobile platform. It is an intelligent monitoring system that determines and controls fruit maturity (GOOD/BAD) by applying OpenCV-based color and shape recognition technology combined with a Backpropagation (BP) multilayer neural network.

The decline in harvesting efficiency due to an aging rural workforce and a shortage of skilled labor. High manufacturing costs of existing harvesting robot systems and low accuracy in determining fruit maturity.

This technology features a wheeled mobile robot equipped with a lift to adjust camera height. It recognizes objects through HSV color space segmentation via OpenCV, binarization, ROI extraction based on moment functions, and Canny edge filtering. It improves recognition accuracy by finalizing maturity assessments using red ratio analysis and a multilayer neural network (BP neural network).

‍

‍

Key Features:

• A main body equipped with mobile wheels, a power supply unit, a communication unit, and a camera unit for capturing the monitoring target
• Temperature/humidity sensors for measuring the environment of the target location and a lift support unit that provides elevation functionality to the camera unit
• A drive unit that raises or lowers the camera unit connected to the lift's movement axis or operates the mobile wheels of the main body
• A control unit that extracts the region of interest from the binarized object image by adjusting HSV values and determines the ripeness level based on the red ratio

‍