Human Vision vs. Computer Vision: A Readability Study in a Virtual Reality Environment

• Published in: 2022 IEEE Conference on Virtual Reality and 3D User Interfaces Abstracts and Workshops (VRW) pp. 61 - 64
• Main Authors: Qing, Zhu, Edara, Praveen
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.03.2022
• Subjects: Applied computing-Operations research; Cloud computing; Computer vision; Conferences; Design methodology; Evaluation methods; Human factors; Human-centered computing-Visualization-Visualization design and evaluation methods; Road sign readability; Text readability; Three-dimensional displays; Transportation; User study; Virtual reality; Visualization
• DOI: 10.1109/VRW55335.2022.00023

Text plays an important role in conveying information to users in a virtual reality (VR) environment. Both VR software and hardware are evolving rapidly to improve text display quality. However, evaluation of text readability still relies on human participants. In this study, cloud computer vision was used to evaluate text readability in VR. Human subjects were recruited to test the same text scenarios. The cloud computer vision-based approach produced results that were consistent with human vision-based recognition. The use of computer vision to automate text readability evaluation could significantly reduce the overall effort and time in developing readable text in VR.