Automatic projection positioning based on surface suitability

Projectors most likely will be embedded into the next generation of mobile devices, and thus projecting pictures and videos into the physical world will be an important use case. However, the projection positioning in the real world is challenged by surface reflections or objects in the environment. As adjusting the projected image manually is cumbersome, in this work, we introduce a camera-based algorithm to measure the projection suitability of arbitrary surfaces by using three classifiers that can be derived automatically and in real-time from a camera image. Through a user study, in that participants rated the projection quality of different surfaces, we developed guidelines to indicate suitable projection spots using the three classifiers. As a proof of concept, we implemented a mobile prototype applying the proposed guidelines. This prototype automatically scales and places projected content at the position in everyday environments that promises the best projection quality.