Overview
The Yahboom K210 Vision Recognition Module is an AI vision recognition module for embedded image processing and local visual inference. It is based on the Kendryte K210 (dual-core 64-bit RISC-V) with about 1TOPS equivalent computing power, and integrates a serial port interface for connecting to controllers such as STM32, Raspberry Pi Pico, Arduino, Micro:bit, MSPM0, and PC.
Hardware resources include a 2MP camera (OV2640), a built-in 2.0-inch LCD capacitive touch screen (320*240), programmable RGB lights, buttons, and an onboard TF card slot (TF card optional) for storing data and model files. It supports open-source MicroPython code, one-click training, and custom model training/export in kmodel format.
Key Features
- Kendryte K210 AI chip: dual-core 64-bit RISC-V; about 1TOPS equivalent computing power; built-in hardware acceleration units (KPU/FPU/FFT)
- 2MP HD camera (OV2640) for AI visual recognition tasks
- 2.0-inch LCD capacitive touch screen, 320*240 resolution
- Serial communication: transmit recognition results/data to external controllers via serial port
- AI visual functions shown: QR code recognition, barcode recognition, AprilTag recognition, handwritten digit recognition, license plate recognition
- Human-related functions shown: face detection, facial feature detection, mask recognition, face recognition, human body recognition, palm recognition (face in demos is blurred)
- Autonomous learning: object recognition, self-learning classification, self-training model (cloud training platform support shown)
- Color recognition: multiple-colors recognition; visual line tracking for line patrol scenarios
- TF card support (package options shown: with or without TF card; poor-quality TF cards may be incompatible)
- MicroUSB data connection shown for real-time firmware upgrade
- Mounting support: fixed holes; compatible with Lego brick holes; optional mounting brackets and 2DOF Electric PTZ accessory shown
Specifications
| Chip | Kendryte K210 |
| Process | TSMC 28nm |
| CPU | Dual-core 64-bit RISC-V |
| Equivalent computing power | About 1TOPS |
| Main frequency | Maximum 400MHz |
| Camera | 2MP HD camera, OV2640 |
| Display | 2.0-inch LCD capacitive touch screen |
| Display resolution | 320*240 |
| Recognition frame rate | QVGA@60fps; VGA@30fps |
| On board interface | Serial port interface |
| Network connection | Not support |
| Model training | Support |
| AI visual recognition cases | 16 AI visual recognition cases |
| Power consumption / working current | 0.8W / 0.16A |
| Size | 57.3*41.2*17.4mm |
| Storage | On board TF card slot; optional TF card for storing data and model files |
What's Included
- K210 vision module
- Package options shown: Without TF card (users prepare TF card and card reader) or With TF card (shipping list includes a TF card + card reader)
Applications
- Embedded AI vision projects and intelligent hardware prototypes
- Robots and smart cars requiring serial output of detection/recognition results
- Color line tracking and color block recognition scenarios
- Optional mounting use: fixed bracket, angle adjustable bracket, heightened angle adjustable bracket (1.2-9.5cm), and 2DOF Electric PTZ applications shown
For product selection advice, integration support, or after-sales help, contact https://rcdrone.top/ or email support@rcdrone.top.
Tutorials
Yahboom K210 Visual Recognition Module Tutorials
Details

Bring edge AI vision to embedded projects with a Kendryte K210 module that combines a 2MP camera, 2.0-inch touch display, and UART output.

The compact board integrates the camera, touch LCD, and onboard controls for quick prototyping and demos.

Choose a package with or without a TF card, and add brackets or a 2DOF pan-tilt accessory for robotics and fixed installs.

Kendryte K210 delivers fast local inference for vision tasks using built-in acceleration (KPU/FPU/FFT) while keeping power low.


A quick comparison helps choose between K210 and higher-performance options based on interfaces, display size, and supported use cases.



Built-in demos cover common machine-vision tasks like QR/barcode scanning, AprilTag detection, and character recognition.

Human-related demos include detection and tracking features, with results delivered over the serial port to an external controller.

Use autonomous learning for custom objects, then extend projects with multi-color recognition and line-following capabilities.

A guided workflow supports one-click training and exporting models in kmodel format for deployment on the module.

The capacitive touch LCD supports interactive UIs, enabling on-device controls, status views, and simple dashboards.

Key hardware features include RGB LEDs, buttons, MicroUSB firmware upgrade, TF card storage, and mounting holes compatible with Lego-style builds.

Connect to popular controllers such as STM32, Raspberry Pi Pico, Arduino, and Micro:bit to send recognition results via UART.


Example projects range from QR-controlled cars and label tracking to road-sign recognition and autonomous driving experiments.

A labeled board map makes it easy to locate the camera, TF slot, MicroUSB, buttons, RGB LEDs, and the external serial port header.

Yahboom provides an online tutorial guide for the K210 AI camera module covering MaixPy setup, LCD and camera use, and UART communication.

The Yahboom K210 vision module comes with organized tutorial documents and MicroPython example code covering basic, AI vision, and communication modules.

The K210 course learning route covers chip basics, development environment setup, experimental cases, GUI operation, AI vision use cases, and expanded applications.

The Yahboom K210 vision module and its mounting brackets are provided with detailed millimeter dimensions to help plan installation and fit.

Angle-adjustable bracket and 20DF electric platform dimensions help plan mounting hole spacing and overall fit during installation.

K210 chip parameters list key details like 400MHz (up to 600MHz) clock speed, 8MB SRAM, 28nm process, and common interfaces including UART, SPI, and I2C.

The K210 vision recognition module kit includes the board, micro USB cable, connection leads, and optional TF card, card reader, and mounting bracket depending on the package.

The K210 vision recognition module kit includes the main module plus wiring cables, TF card and reader, and bracket or PTZ mounting options for setup.

The K210 vision module pairs an OV2640 camera with a compact 2.0-inch touch LCD for on-device viewing and control.
Related Collections
