RGB Classification Determination with Different Light Intensity Using Pixy CMUcam5

The surrounding conditions may influence the robustness of tracking performance. The objective of this research is to get the best tracking performance using Pixy CMUcam5 tracking camera while tracking the colour pattern. The RGB values from PixyMon image frame-grabber were extracted from the different light intensity. Each of the information taken from Pixy CMUcam5 was tabulated into the table. The result shows that the most suitable conditions for Pixy CMUcam5 to operates is within 58 to 205 lx of light in a closed environment as it gives a smooth and uniform RGB value. It shows that the camera may have too much of tracking errors if the light is too dim or too high in intensity. Several information such as RGB boundaries and RGB average values were measured with several algorithms using MATLAB. The tracking error rate being computed to get the best conditions for tracking performance. As the conclusion, the determinations of best tracking performance for CMUcam5 is important and it can be applied to other systems like mobile robot tracking.