Feather Damage Monitoring System Using RGB-Depth-Thermal Model for Chickens.

Area of Science:

• Animal Science
• Biotechnology
• Agricultural Engineering

Background:

• Feather damage is a significant welfare issue in laying hens, impacting animal health.
• Infrared thermography can detect surface temperature changes indicative of inflammation and dermal damage.
• Objective assessment of feather damage depth is crucial for effective management.

Purpose of the Study:

• To develop an integrated approach for assessing feather damage in laying hens.
• To combine visible light and infrared thermography for enhanced damage evaluation.
• To create a 3D model incorporating RGB-Depth-Thermal data for precise damage assessment.

Main Methods:

• A novel pipeline was developed to generate RGB-Depth-Thermal maps of chickens.
• Stereo matching of binocular color images provided depth information.
• Heterogeneous image registration aligned thermal infrared images with color and depth data.
• Deep learning-based segmentation isolated the chicken from the background.
• A 3D chicken model was reconstructed using color, depth, thermal, and mask images.

Main Results:

• The proposed RGB-Depth-Thermal model enabled accurate reconstruction of chicken surface data.
• The fused data approach improved the assessment of feather damage depth compared to 2D methods.
• The system demonstrated effectiveness in evaluating feather damage during both day and night.

Conclusions:

• The developed RGB-Depth-Thermal mapping pipeline provides an objective tool for assessing feather damage in poultry.
• This method offers a valuable reference for advancing poultry farming practices and animal welfare.
• Integrating multiple imaging modalities enhances the capability to evaluate sub-dermal poultry health issues.