Embedded System Object Tracking using CMU Camera

Computer vision has gone long way from 1980s till now. Before, computer vision generally so robust and expensive. However, due to advancement in embedded processing, computer vision can be implemented in cheaper computer system. Raspberry Pi and CMUcam4 provides fantastic platform for interfacing and also provides a simple robotic object motion tracking at school-level or university-level. These simple computer vision systems can benefits in many ways such as being applied to police task force, or bomb squad purpose, or firefighting purpose. To achieve the interfacing, serial communication between the RPi and CMUcam need to be achieved. Then we’ll develop the color tracking algorithm to track a color and ask the camera to follow that colored object movement. From the troubleshooting and findings, interfacing RPi and CMUcam requires a lot of tweaking as both consists of two different systems. In the end, only servo platform does not successfully work.