Tracking systems are a widely used technology in physical computing. Without a positioning system, many applications like robotic systems could not function properly. Especially Virtual Reality / Augmented Reality headsets rely heavily on the position data of such systems. For this emerging technology, accurate and fast position data is crucial, in order to prevent side effects like motion sickness. Furthermore, tracking systems have to be cheap and portable to be competitive on the mass market. In this thesis, a new kind of tracking system is presented, based on the refraction of light and optical focus. The main part of the system forms a linear image sensor together with a 4-piece lens system. Light originates from Light Emitting Diodes (LEDs) placed in the tracking environment. These LEDs are controlled by a microcontroller one at a time. The LED light cones need to be overlapping, in order to get at least two intensity measurements by different light sources. Light rays travel through the lens system, which focuses light on the sensor. When the sensor moves, the angle of the refracted light changes, thus reaching a different part of the sensor. A microcontroller can then compute the tracking systems change of position. The project can be split into two parts: The first part consists of a tracking system simulation. The simulation makes use of Ray Transfer Matrix Analysis, a ray tracing technique utilized in the design of optical systems. Using the simulation, the size of the tracking area has been calculated and the right lenses with correct distances to each other have been chosen. The second part consists of the actual implementation. A functional prototype has been developed, to turn the insights of the simulation into a real-life solution. A microcontroller is used to read the linear image sensor and control the LEDs. The intensity values are sent to a Raspberry Pi, which calculates the position change of the tracker. With only one linear image sensor, the system is able to track translations in two dimensions. However, through adding an additional sensor with corresponding LEDs, tracking of a third axis is possible in theory. Finally, the accuracy and latency of the tracking system prototype have been evaluated.