“Study of Augmented Reality Methods for Real Time Recognition and Tracking of Untextured 3D Models in Monocular Images”

The paper was published by the 10th IEEE International Symposium on Mixed and Augmented Reality (ISMAR 2011). The paper proposes a new real time augmented reality tool to aid maintenance operations by giving workers virtual instructions that allow them to perform disassembly tasks more efficiently.
The prototype is a system that is able to automatically generate all the necessary data from object geometries. All the operations to be carried out in the disassembly task as well as the visual characteristics that allow the objects to be recognized in the image are automatically extracted in a pre-processing stage, thereby eliminating the exhausting and costly need for user intervention.
Furthermore, the system runs in real time and uses common, low cost hardware components: a web camera to capture images of the scene; a PC to process images and make the calculations needed to present the virtual instructions; and a monitor to display the results.
Although the prototype is geared toward industrial applications, it has other practical applications as well. For example, it can be used to facilitate furniture assembly/disassembly by giving easier to understand and more detailed information than in traditionally provided in paper instructions.