The Pixy CMUcam5 Image Sensor is an ultra-fast vision sensor you can quickly "teach" to find objects, and it connects directly to Arduino and other controllers. It’s great for robotics or any other application that requires computer vision. Pixy pairs a powerful dedicated processor with the image sensor. Pixy processes images from the image sensor and only sends the useful information (e.g. black cat detected at x=64, y=113) to your microcontroller. And it does this at frame rate (50 Hz). The information is available through one of several interfaces: UART serial, SPI, I2C, USB, or digital/analogue output. So you’re Arduino or other microcontroller can talk easily with Pixy and still have plenty of CPU available for other tasks.
The CMUcam5 Pixy uses a "colour filter" algorithm based on "colours" to detect objects. Most of us are familiar with the concepts RGB (red, green and blue) to represent the colours. The Pixy CMUcam5 calculates the hue and saturation of each pixel RGB sensor from the video and used as a primary filter settings. If in the majority of cases, the hue of an object remains unchanged due to changes in lighting and exposure, this phenomenon may however work against the filtering algorithm.