Method to quantify perceived scene movement velocity in virtual environment

Abstract

There is a significant difference between the perceived speed and the actual speed of a moving object. In most researches concerning speed perception, either ‘point’ or ‘line’ is used in the experiments. There have been few studies that have utilized a two-dimensional object in their experiments related to speed perception. In this paper, the perceived speeds of a linearly moving object and an expanding object were measured using the Magnitude Estimation method. In the case of linear movements, participants overestimated the speed of an object when the actual speed did not exceed the standard speed. Additionally, they underestimated the speed when the actual speed was higher then the standard speed. For expanding movements, participants underestimated the speed of an object regardless of the actual speed, and the more the actual speed differed from the standard speed, the more it was underestimated. Finally, for both a linearly moving object and an expanding object, the perceived speed was different according to the size of an object. The perceived speed decreased as the size of the object increased. This paper proposes an objective measure to quantify the perceived scene movement velocity in a Virtual Environment (VE). This objective measure can be used in various relative research fields including comparing two different VEs and presenting a guideline for cybersickness prevention. The proposed method is a composition of scene complexity and scene movement velocity. While existing researches emphasize the scene complexity, this paper emphasizes the scene movement velocity, and explains Movement Velocity (MV) and Scene Movement Velocity (SMV). The proposed method was verified by applying it to the experiment on the effect of the object size in speed perception. The result showed that perceived speed decreased as the size of the object increased. So did both MV and SMV. A strong correlation was found between perceived speed and MV, SMV. The above results show ...