Automatic detection of animals in mowing operations using thermal cameras

Area of Science:

• Agricultural Engineering
• Wildlife Conservation
• Ecology

Background:

• Modern agricultural practices with high-efficiency equipment increase the risk of wildlife injury or mortality.
• Ground-nesting birds and mammals are particularly vulnerable due to their nesting habits and predator avoidance behaviors.
• Existing wildlife-friendly farming methods often reduce operational efficiency, necessitating technological solutions.

Purpose of the Study:

• To assess the suitability of thermal imaging and digital image processing for automatic detection of wild animals in agricultural settings.
• To evaluate the effectiveness of these technologies in reducing wildlife mortality during farming operations.
• To explore advancements in wildlife-friendly farming through automated detection systems.

Main Methods:

• Utilized thermal imaging combined with digital image processing techniques.
• Conducted tests in grassland habitats to detect study animals (chicken and rabbit).
• Analyzed detection rates under various environmental conditions, including different grass densities.

Main Results:

• Achieved high precision in detecting study animals across different test scenarios.
• Observed a reduced detection rate in areas with the densest grass cover.
• Demonstrated the potential of the integrated system for wildlife detection in agriculture.

Conclusions:

• Thermal imaging and digital image processing are effective tools for automatically detecting wildlife in agricultural environments.
• This technology can significantly contribute to improving wildlife-friendly farming practices.
• Further development could mitigate challenges posed by dense vegetation, enhancing conservation efforts.