Innovation Objective
Current mobile visual search solutions achieve search results based on the appearance of real 3D objects in images captured by mobile devices. These solutions fail if the different real 3D objects appear similar in the captured image or it is from a poster showing a picture of a real 3D object.
To solve above problems, the novel solution will capture and use the underlying 3D geometry of query objects as well as their appearance.
Expected Outcome
A novel mobile search solution that captures and sends robust 3D features from a mobile device / system and that recognizes real 3D objects based on the underlying 3D geometry. This is an advantage over current search solutions that are based on visual appearance only. For example, our new solution will allow autonomous cars to recognize obstacles on the road ahead more accurately, and hence, increase the safety on our future roads.