Identification of granite varieties from colour spectrum data

Area of Science:

• Geoscience
• Materials Science
• Data Science

Background:

• The granite industry in Northwest Spain processes diverse granite types with varying market values.
• Accurate identification and classification are crucial for optimizing stock management and traceability systems.

Purpose of the Study:

• To develop an automated methodology for identifying and classifying granite varieties.
• To leverage spectral information and machine learning for precise granite characterization.

Main Methods:

• Utilizing spectrophotometer to capture spectral information from granite at different processing stages.
• Applying functional machine learning techniques to analyze spectral data for identification.

Main Results:

• Successful development of a methodology for automatic granite variety identification.
• Demonstrated potential for optimizing stock control and traceability in the granite sector.

Conclusions:

• Automated spectral analysis with machine learning offers an effective solution for granite classification.
• The proposed method enhances efficiency and accuracy in the natural stone processing industry.