I, in collaboration with a team, am making an Arduino-controlled robot car which must complete a circuit race. In the circuit race, the boundaries of the track are marked with pylons, thus pylon detection is a must. We are using an Xbox Kinect camera for this project, and have successfully been able to detect the orange pixels of the pylons as well as some background pixels. However, to identify the pylons, we need to further process the data given. We are using pixel grouping for now, but are looking for more efficient and simpler ideas. Thanks for the help in advance!