Computer-Illustrated Reality

Computer-illustrated reality is a novel approach to combine computer-generated technical illustration with the first- person experience of Augmented Reality. Computer-generated technical illustration is concerned with the automatic creation of digital images which mimic traditional illustration techniques found in textbooks and manuals. Previously, such techniques were mostly targeting still image generation or desktop virtual reality. In contrast, Augmented Reality combines a user`s live view of real world objects with computer generated images. A main goal of the project is therefore to bring illustrative depictions into the new medium of Augmented Reality. This is achieved by first acquiring a digital model of the relevant objects in the real world, and then analyzing objects and first-person video images to infer how the illustrations should be rendered. The goal is to match the real scene, but more importantly to convey the important illustrative goals in an efficient and effective manner. We are investigating a number of techniques to achieve this. Firstly, adaptive visualization techniques are required to address the continuously changing nature of first-person Augmented Reality. Secondly, illustrative visualization is concerned with appropriate stylization as well as object visibility. Thirdly, narrative visualization deals with the temporal domain. The research is complemented by quantitative evaluation of the effects the investigated techniques have on real users. This research project has great potential in applications concerned with computer- based learning, training and remote consultation.

Computer-illustrated reality is a novel approach to combine computer-generated technical illustration with the first-person experience of Augmented Reality. Computer-generated technical illustration is concerned with the automatic creation of digital images which mimic traditional illustration techniques found in textbooks and manuals. Previously, such techniques were mostly targeting still image generation or desktop virtual reality. In contrast, Augmented Reality combines a user's live view of real world objects with computer generated images. The goal of the project is to bring illustrative depictions into the new medium of Augmented Reality and thus to maximize the effectiveness of the resulting visualization. We focus on the most commonly used illustrative elements, such as annotations and their layout, X-Ray visualization and illustrative graphics of dynamic processes.We generate computer-illustrated Augmented Reality by first capturing a digital model of relevant real-world objects. Subsequently, we align the digital objects with the live video data of the Augmented Reality system, which allows us to modify the real objects within the live video data in a way that the resulting visualizations possess the characteristics of traditional illustrative graphics. In order to address the continuously changing nature of first-person Augmented Reality, we have developed rendering methods which adapt the properties of the visualization in each frame. This includes an adaptive contrast enhancement in order visually amplify image elements which are important to communicate the shape of the object, and frame coherent label placement techniques to ensure stable label layouts in dynamic Augmented Reality environments.In order to generate illustrative Augmented Reality visualizations of human actions, we have implemented a system which is able to extract motions from traditional two-dimensional illustrations. We use the extracted motions to create three-dimensional animations as well as three-dimensional dynamic glyphs, which we present to the user in Augmented Reality to provide instructions on how to perform certain actions.This research project has great potential in applications concerned with computer-based learning, training and remote consultation. The research done in this project has carried out several software prototypes of Augmented Reality applications using illustrative visualizations.