Image-Based Texture Styling for Motion Effect Rendering
Beomsu Lim, Sangyoon Han, Seungmoon Choi · 2021
A motion platform provides the vestibular stimuli that elicit the sensations of self-motion and thereby improves the immersiveness. A representative example is a 4D Ride, which presents a video of POV shots and motion effects synchronized with the camera motion in the video. Previous research efforts resulted in a few automatic motion effect synthesis algorithms for POV shots. Although effective in generating gross motion effects, they do not consider fine features on the ground, such as a rough or bumpy road. In this paper, we propose an algorithm for styling the gross motion effects using a texture image. Our algorithm transforms a texture image into a high-frequency style motion and merges it with the original motion while respecting both perceptual and device constraints. A user study demonstrated that texture styling could increase immersiveness, realism, and harmony.