AI camera design capable of learning through clicks
Robotics and open source hardware provider DFRobot launched HuskyLens on July 25th, a user-friendly, artificial intelligence vision sensor with various powerful algorithms built in, designed for creative projects through the simple philosophy of click, learn, and play.
Unlike other similar products on the market, HuskyLens has several types of built-in image processing algorithms, allowing creators to start experimenting without understanding such algorithms. Hence, it is not only a time-saving device for experienced makers, but an easy-to-use powerful tool for enthusiastic future makers as well.
With pre-installed algorithms, HuskyLen is equipped with facial recognition, object recognition, object tracking, line following, colour detection, and tag detection. Users can also switch between algorithms with one single click. Nevertheless, algorithms like line-following and tag recognition can be activated simultaneously, for instance, once applied to a small robot, one can not only follow a line, but can also determine in which direction to turn.
What’s more, HuskyLens can improve its abilities through learning. With one single click, the device can learn new objects, faces, and colours. Moreover, by holding the button down, HuskyLens can continually learn the target, even from different perspectives and in various ranges. Additionally, the AI camera carries a two inch inch display, which means users don’t have to connect to a PC for monitoring, what you see is what you get!
Ricky Ye, CEO of DFRobot said: “HuskyLens has been designed with creators and makers in mind and has infused the latest AI technology with compact design and exceptionally fast processors to ensure that users are only limited by their imagination.
“Our development team has worked incredibly hard to develop such a fantastic product and I look forward to seeing the different ways that HuskyLens is utilised by our community of innovators.”
HuskyLens is compatible with a range of microcontrollers including Arduino, Raspberry Pi, LattePanda, and micro:bit, and is now available for backing on KickStarter.