Large-scale quantitative analysis of painting arts

Area of Science:

• Art History
• Computer Vision
• Digital Image Processing

Background:

• Understanding the aesthetic qualities of painting art has been a long-standing scientific endeavor.
• Advancements in digital image acquisition enable large-scale statistical analysis of artistic paintings, bridging art and science.
• Quantitative analysis offers new perspectives on art historical trends and techniques.

Purpose of the Study:

• To quantitatively analyze artistic paintings using digital image processing techniques.
• To investigate color usage, color variety, and brightness roughness in paintings.
• To establish correlations between quantitative image measures and art historical periods/techniques.

Main Methods:

• Utilized digital image processing techniques for quantitative analysis.
• Investigated three key image measures: individual color usage, color variety, and brightness roughness.
• Analyzed a large-scale database of artistic paints.

Main Results:

• Observed distinct differences in color usage between classical paintings and photographs.
• Identified significantly lower color variety in medieval period paintings.
• Found that the increment of the roughness exponent correlates with the advancement of painting techniques like chiaroscuro and sfumato, aligning with historical contexts.

Conclusions:

• Digital image processing provides a quantitative framework for analyzing art historical data.
• Color usage and variety are significant indicators of different art periods and styles.
• Brightness roughness analysis can quantitatively reflect the evolution of painting techniques throughout history.