Photogrammetric Co-Processing of Thermal Infrared Images and RGB Images

Area of Science:

• Photogrammetry
• Thermal Infrared Imaging
• Computer Vision

Background:

• Spatial orientation of thermal infrared (TIR) data is crucial for applications.
• Photogrammetric processing of TIR images is challenging due to low resolution and lack of visible features.
• Existing methods struggle to accurately integrate thermal data into 2D and 3D spatial contexts.

Purpose of the Study:

• To develop and present novel methods for the photogrammetric co-processing of TIR and RGB images.
• To overcome the inherent limitations of processing TIR images for spatial analysis.
• To enable the creation of augmented 2D and 3D results with accurate thermal information.

Main Methods:

• Designed two methods for photogrammetric co-processing of TIR and RGB images.
• Required a fixed camera system for simultaneous TIR and RGB image capture.
• Developed 'sharpening' (for augmented orthophotos and 3D model textures) and 'reprojection' (for thermal-augmented point clouds) techniques.

Main Results:

• Successfully co-processed TIR and RGB images, overcoming resolution and feature limitations.
• Achieved accurate spatial localization, scaling, and measurement of thermal occurrences.
• Generated augmented orthophotos, 3D model textures, and thermal-augmented point clouds.

Conclusions:

• The proposed co-processing methods effectively integrate thermal infrared information into photogrammetric products.
• These techniques enhance the utility of thermography for detailed spatial analysis and measurement.
• The methods provide valuable tools for applications requiring precise thermal data orientation and quantification.