
1 / 7 · Click to enlarge
AI Summary
A self-contained AI vision sensor with a 2.0-inch IPS screen that recognises faces, objects, lines, colours and tags without coding. Press the learning button to teach it new things and see results instantly on the display.
Key Highlights
- Kendryte K210 processor with OV2640 2.0 MP camera for on-board AI vision
- Built-in face, object, line, colour and tag recognition algorithms
- 2.0-inch IPS screen (320 x 240) for live view and parameter tuning
- Simple function and learning buttons for no-code operation
- I2C and UART interfaces for easy connection to microcontrollers
- Wide 3.3 to 5 VDC input with low current draw
Ideal For
- Makers adding vision to Arduino or Raspberry Pi projects
- Robotics enthusiasts building line-following or object-tracking robots
- STEM educators teaching AI and computer vision concepts
- Students working on interactive installations or smart devices
- Prototypers who need quick vision feedback without coding
Buying Tip
Check that your host board can supply the required current (up to 320 mA) and supports I2C or UART before purchasing.
Elecvora expert review
Our verdict
The DFRobot HuskyLens is a remarkably self-contained AI vision sensor that removes the usual barriers to adding machine vision to electronics projects. Its on-board display and button-based learning make it genuinely usable without a computer, though it is best suited to projects that need one of its built-in algorithms rather than custom vision models.
AI-assisted analysis based on the manufacturer's specifications. Not a hands-on lab test.
Pros
- No coding required for basic setup and teaching
- Built-in screen simplifies tuning and debugging
- Multiple recognition modes in one compact module
- Standard I2C and UART interfaces for easy integration
- Wide voltage range and low current draw
Cons
- Limited to the pre-installed algorithms
- Requires a host microcontroller for advanced control
- No enclosure or case included
Best for
Makers and educators who want quick, code-free vision capabilities
Not ideal for
Developers needing custom neural network models or high-resolution imaging
Specifications
| Item Type | Smart Vision Sensors |
|---|---|
| Processor | Kendryte K210 |
| Image Sensor | OV2640 (2.0 Megapixel Camera) |
| Rating Voltage (VDC) | 3.3 to 5 |
| Current consumption | [email protected], [email protected] |
| Interface Type | I2C, UART |
| Display | 2.0-inch IPS screen |
| Resolution | 320 x 240 px |
| Length (mm) | 52 |
| Width (mm) | 44.5 |
| Shipping Weight | 0.1 kg |
| Shipping Dimensions | 9 × 5 × 3 cm |
What's in the box
- 1 x HuskyLens Mainboard
- 6 x M3 Screws
- 6 x M3 Nuts
- 1 x Small Mounting Bracket
- 1 x Heightening Bracket
- 1 x Gravity 4-Pin Sensor Cable
Frequently Asked Questions
Do I need to program HuskyLens to use it?
What is the power requirement for HuskyLens?
What comes in the package?
Can HuskyLens recognise objects it has not seen before?
What interfaces does HuskyLens support?
How do I care for the HuskyLens?
Overview
The DFRobot HuskyLens is a smart vision sensor designed to make machine vision accessible. At its core is the Kendryte K210 processor paired with an OV2640 2.0-megapixel image sensor, giving it the ability to run recognition algorithms entirely on-board. A 2.0-inch IPS display with 320 x 240 resolution shows what the camera sees and overlays detection results, so you can work directly on the module without a computer.
HuskyLens is operated using two buttons: a function button to cycle through algorithms and a learning button to teach it new objects. This no-code approach means you can focus on building your project rather than programming vision routines.
Key Features
- Multiple vision functions: face recognition, object tracking, object recognition, line tracking, colour recognition and tag (QR code) recognition are built in.
- On-board display: the 2.0-inch IPS screen provides immediate visual feedback for setup and debugging.
- Simple teaching: press the learning button to register a new object; long-press to add more examples from different angles and distances for improved accuracy.
- Flexible connectivity: communicates over I2C or UART, making it compatible with a wide range of microcontrollers and development boards.
- Compact and lightweight: measures 52 mm by 44.5 mm and weighs approximately 0.1 kg for shipping, so it suits small robots and portable devices.
Specifications Explained
The Kendryte K210 is a dedicated AI processor that handles the recognition tasks locally, so no cloud connection is needed. The OV2640 camera captures 2.0-megapixel images, which is sufficient for the built-in algorithms. The 2.0-inch IPS screen offers a clear, colourful view for framing and reviewing detections. Power requirements are modest: the module accepts 3.3 to 5 VDC and draws about 320 mA at 3.3 V or 230 mA at 5 V, so it can be powered from common development boards or battery packs. I2C and UART interfaces give you flexibility in how you integrate the sensor with your control electronics.
Compatibility and Requirements
HuskyLens connects to host devices via I2C or UART. It requires a power supply within the 3.3 to 5 VDC range and draws up to 320 mA. No specific development platforms are listed in the product data, but the included Gravity 4-pin sensor cable is provided for connection. The module can be used standalone with its buttons and screen, or connected to a microcontroller for programmatic control.
Typical Uses
- Adding face detection or object tracking to robotics projects
- Line-following robots and colour-sorting machines
- Interactive installations that respond to tags or colours
- Educational demonstrations of machine learning and computer vision
- Prototyping smart vision features without writing complex code
What's in the Box
- 1 x HuskyLens mainboard
- 6 x M3 screws
- 6 x M3 nuts
- 1 x small mounting bracket
- 1 x heightening bracket
- 1 x Gravity 4-pin sensor cable
Quick specs
- Item Type
- Smart Vision Sensors
- Processor
- Kendryte K210
- Image Sensor
- OV2640 (2.0 Megapixel Camera)
- Rating Voltage (VDC)
- 3.3 to 5
- Current consumption
- [email protected], [email protected]
- Interface Type
- I2C, UART



















