Looking Through Paintings by Combining Hyper-Spectral Imaging and Pulse-Compression Thermography

Area of Science:

• Art conservation science
• Non-destructive testing
• Multispectral imaging

Background:

• Artwork inspection requires identifying defects before restoration.
• Non-destructive testing methods are crucial for preserving multi-layered art structures.
• Traditional methods may not reveal all surface and subsurface anomalies.

Purpose of the Study:

• To evaluate infrared thermography and hyperspectral imaging for artwork defect detection.
• To identify both fabricated and non-fabricated targets in a canvas painting.
• To enhance defect identification using statistical post-processing and data fusion.

Main Methods:

• Applied pulse-compression infrared thermography with modulated LED heat sources.
• Utilized hyperspectral imaging to detect surface and subsurface features.
• Employed principal component and independent component analyses for target retrieval.
• Fused data from both techniques for comprehensive analysis.

Main Results:

• Both techniques successfully identified various targets in the mock artwork.
• Statistical algorithms significantly increased the number of detected anomalies.
• Data fusion provided a clear map integrating surface, subsurface, and deeper information.
• Minimizing heat radiation was achieved using modulated LED chips.

Conclusions:

• Infrared thermography and hyperspectral imaging are valuable tools for art inspection.
• Statistical analysis and data fusion enhance the detection capabilities.
• Combined methods offer a comprehensive approach to identifying artwork defects for preservation.