Assessment of newly designed fonts for visual accessibility

Area of Science:

• Visual perception and typography.
• Human-computer interaction.
• Accessibility research.

Background:

• Font design aims to maximize text accessibility for diverse users, including those with low vision.
• Evaluating font visual accessibility involves assessing reading efficiency across various print sizes.
• New fonts require rigorous testing to ensure usability for all individuals.

Purpose of the Study:

• To compare behavioral and automated methods for assessing font visual accessibility.
• To evaluate the accessibility of the newly designed font, ACT Easy, for normal and simulated low vision.
• To determine if automated methods can replicate human performance in font accessibility evaluation.

Main Methods:

• Experiment 1: Behavioral (psychophysical) testing with 22 normally sighted adults using the MNREAD computerized test.
• Experiment 1: Assessed reading acuity, critical print size, and reading speed under normal and simulated low-vision (20/90 acuity) conditions.
• Experiment 2: Automated evaluation using eleven Optical Character Recognition (OCR) models to estimate reading acuity across different fonts.

Main Results:

• ACT Easy Regular font performed comparably to Courier in reading acuity and critical print size for both viewing conditions.
• ACT Easy Regular and Gotham fonts were preferred by participants in ranking.
• Nine of eleven OCR models demonstrated human-like changes in reading acuity across viewing conditions and fonts.

Conclusions:

• Both behavioral and automated methods can identify differences in font visual accessibility.
• Automated methods using OCR models offer a labor-efficient alternative to human testing for font evaluation.
• Font accessibility research can benefit from integrating automated evaluation techniques.