Gravity: HUSKYLENS with Silicone Case

The Gravity: HUSKYLENS is an easy-to-use AI machine vision sensor equipped with multiple built-in algorithms, including face, object, colour, tag, and line recognition. Fitted with an on-board IPS display and a protective silicone case, it is ideal for students, educators, and makers adding computer vision to micro:bit, Arduino, and Raspberry Pi projects without writing complex code.

Built-in machine vision algorithms

The sensor comes pre-loaded with algorithms for face recognition, object tracking, object recognition, line tracking, colour recognition, and tag recognition. These built-in functions allow makers to implement vision-based tasks without dealing with complex neural network coding.

Broad development board compatibility

Featuring standard UART and I2C connectivity, the module interfaces smoothly with micro:bit, Arduino, Raspberry Pi, and LattePanda. The included Gravity 4-pin sensor cable allows for quick wiring to compatible expansion boards.

Integrated display and on-device tuning

A built-in 2.0-inch IPS screen with 320x240 resolution provides real-time visual feedback and on-device parameter configuration. Users can switch algorithms with the function button and trigger learning modes instantly without connecting to a computer.

Continuous multi-angle learning

The on-board machine learning technology allows the sensor to learn new faces and objects directly from the environment. Long-pressing the learning button lets the camera continuously train across various angles and distances for improved accuracy.

High-speed Kendryte K210 processor

Powered by a dedicated Kendryte K210 AI processor, the sensor runs neural network algorithms up to 1,000 times faster than an STM32H743 chip. Paired with a 2.0-megapixel OV2640 camera, this processing power enables rapid detection of fast-moving targets.

Hand gesture position tracking

HUSKYLENS can be trained to recognize custom hand movement patterns and feed live coordinate data to your controller. This capability simplifies the creation of interactive gesture-controlled systems, smart toys, and access control projects.

Calibration-free line following

Line tracking is activated with a single click of the learning button, bypassing the need for tedious manual parameter adjustments. This enables autonomous mobile robots to quickly learn, detect, and follow lines in changing environments.

Video

Downloads & documentation

• Product wiki (DFRobot)