Amano, ToshiyukiKagawa, TaichiOrlosky, Jason and Reiners, Dirk and Weyers, Benjamin2021-09-072021-09-072021978-3-03868-142-71727-530Xhttps://doi.org/10.2312/egve.20211327https://diglib.eg.org:443/handle/10.2312/egve20211327In this study, we propose a projection registration method for the projections from a continuously moving vehicle for driver vision assistance during night driving. Accordingly, we employ a context-aware projection technique with adaptive pixel map- ping generation. Because vehicle movements lead to misalignment of the projection latency, a co-axial projector-camera con- figuration or high frame rate processing cannot solve this problem. However, adaptive pixel mapping corrects pixel mapping according to the vehicle speed and achieves a misalignment-free dynamic projection mapping. The effectiveness of the proposed method was evaluated through experiments using a moving projector-camera system mounted on a motorized linear stage.Human centered computingMixed / augmented realityComputing methodologiesMixed / augmented realityProjection Alignment Correction for In-Vehicle Projector-Camera System10.2312/egve.2021132747-51