Modeling of Moving Objects with RGB-D Cameras

Abstract

The purpose of this project is to create a three-dimensional model of an object using an RGB-D sensor. The RGB-D camera used in this project is the Microsoft Kinect 2.0; as it is relatively cheap, easy to obtain for students and provides fairly accurate results. The Kinect measures the depth by using by projecting infrared (IR) dots with a known pattern and using an IR receiver to measure the depth of the object based on the deviation of these projections. Those depth images could be represented as point clouds which could later be merged together to create a more accurate 3D representation. Microsoft have released their own source codes in their software development kit (SDK) which includes Kinect Fusion. Kinect Fusion allows Kinect users to create a 3D model of an object. The source codes will be used as a reference and as a starting point for the project. A combination of C# and MATLAB will be used in order to reduce errors arising in the 3D model as much as possible.