The Pixy CMUcam5 is a remarkably fast image sensor that is able to be taught to find objects of various shapes, sizes and colors. With the capability of tracking hundreds of objects simultaneously, the Pixy can be used in robotics, home automation, colour coding and more — all the while only providing you with the information that you actually want. Pixy really is a fast and easy-to-use vision system equipped with a dedicated NXP LPC4330 processor, making it capable of handling large loads of data from its image sensor and processing it quickly while outputting what it detects 50 times per second.
Pixy processes images from the image sensor and only sends the useful information to your microcontroller, doing so at 50 frames per second. This means each Pixy CMUcam5 processes an entire 640x400 image frame every 1/50th of a second (20 milliseconds). The information is available through one of several interfaces: UART serial, SPI, I2C, USB or digital/analogue output. So your Arduino, Raspberry Pi, BeagleBone or other microcontroller can easily talk with Pixy and still have plenty of CPU available for other tasks you need them to do.
Pixy uses a colour-based filtering algorithm to detect objects. Colour-based filtering methods are popular because they are fast, efficient and relatively robust. Pixy calculates hue and saturation of each RGB pixel from the image sensor and uses these as the primary filtering parameters. The hue of an object remains largely unchanged with changes in lighting and exposure. Changes in lighting and exposure can have a frustrating effect on colour filtering algorithms, causing them to break. Pixy’s filtering algorithm is robust when it comes to lighting and exposure changes.