A Semi-Automated RGB-Based Method for Wildlife Crop Damage Detection Using QGIS-Integrated UAV Workflow

Area of Science:

• Agricultural Science
• Remote Sensing
• Wildlife Management

Background:

• Wildlife damage to crops, especially large monocultures like maize, poses significant agricultural challenges.
• Accurate and scalable monitoring methods are crucial for effective crop management and wildlife impact assessment.

Purpose of the Study:

• To develop and validate a semi-automated process for detecting wildlife-induced crop damage using Unmanned Aerial Vehicle (UAV)-acquired RGB imagery.
• To create a user-friendly, QGIS-integrated tool for non-specialist users to assess crop damage.

Main Methods:

• Calculation of vegetation indices (ExG, GLI, MGRVI) from standardized UAV RGB orthomosaics.
• Application of k-means clustering to classify vegetation vigor into five classes.
• Semi-automated damage classification with interactive threshold adjustment via a QGIS plugin for Drone Data Analysts (DDAs).

Main Results:

• The method achieved 81% overall accuracy in identifying 7 hectares of damage in a 50-hectare maize field.
• Damage was most concentrated in areas with moderate and low vegetation vigor.
• The approach proved effective and scalable for crop damage assessment.

Conclusions:

• The proposed RGB-based, QGIS-integrated method is a repeatable, cost-effective, and field-operable alternative for crop damage assessment.
• This technique supports precision agriculture practices and wildlife population management by providing accurate damage data.