The goal of the Carnegie Mellon's CMUcam project is to provide simple vision capabilities to small-embedded systems in the form of an intelligent sensor. The CMUcam open source programmable embedded color vision sensors are low-cost low-power sensors for mobile robots. You can use the CMUcam vision systems to do many different kinds of on-board real-time vision processing tasks. The CMUcam4 is a fully programmable embedded computer vision sensor. The main processor is the Parallax P8X32A (Propeller Chip) connected to an OmniVision 9665 CMOS camera sensor module.
The CMUcam4 can be used to track colors or collect basic image statistics. The best performance can be achieved when there are highly contrasting and intense colors. For instance it can easily track a red ball on a white background but it would be hard to differentiate between different shades of brown in changing light. Tracking colorful objects can be used to localize landmark follow lines or chase moving beacons. Using color statistics it is possible for the CMUcam4 to monitor a scence detect a specific color or do primitive motion detection. If the CMUcam4 detects a drastic color change then chances are something in the scene changed. Using "Line Mode" the CMUcam4 can generate low resolution binary images of colorful objects. This can be used to do more sophiscated image processing that includes line following with branch detection or even simple shape recognition. These more advanced operations require custom algorithms to post process the binary iages sent from the CMUcam4. As is the case with a normal digital camera this type of processing might require a computer or at least a fast microcontroller.