A toolbox for calculating quantitative image properties in aesthetics research.

Area of Science:

• Visual Aesthetics
• Computational Imaging
• Psychophysics

Background:

• Quantitative image properties are crucial for understanding visual aesthetic appreciation.
• Inconsistent methodologies and tools hinder comparability in visual aesthetics research.
• Lack of standardized, accessible tools limits reproducible research in image aesthetics.

Purpose of the Study:

• To introduce the Aesthetics Toolbox, an open-access software for calculating quantitative image properties.
• To enhance the accessibility and comparability of research findings in visual experimental aesthetics.
• To provide researchers with a unified and simplified method for image property analysis.

Main Methods:

• Developed an open-access, user-friendly toolbox with a click-and-drop web interface.
• Integrated and translated original scripts from four research groups into Python 3.
• Included a comprehensive set of image properties: dimensions, color/lightness statistics, complexity, symmetry, balance, Fourier spectrum, fractal dimension, self-similarity, entropy, and CNN-based variances.
• Ensured consistency between original and Python-translated scripts.

Main Results:

• The Aesthetics Toolbox offers a standardized set of quantitative image properties.
• The toolbox is compatible with most devices and features an intuitive web interface.
• Original research scripts were successfully integrated and validated in Python 3, ensuring result consistency.
• The tool facilitates easier and more comparable analysis of visual aesthetics research data.

Conclusions:

• The Aesthetics Toolbox provides a standardized and simplified solution for calculating image properties in visual aesthetics research.
• This open-access resource promotes greater accessibility and comparability of scientific results.
• The toolbox supports reproducible research by offering validated, consistent image analysis methods.