Yahboom K230 AI Development Board Instruction Manual

1. Introduction to the Yahboom K230 AI Development Board

The Yahboom K230 is an advanced AI development board designed for visual recognition and embedded applications. It features a 1.6GHz high-performance chip, a 2.4-inch display, and supports various AI visual recognition functions. This module is suitable for education, maker projects, and industrial visual development.

Image: Diagram illustrating the K230 AI computing platform chip.

The K230 module offers significant computing power, with its KPU being 13.7 times more powerful and its CPU 8.5 times more powerful than the K210 visual module. This enables real-time operation of complex AI models for tasks such as image recognition and voice processing.

Image: Technical positioning of the K230 Vision Recognition Module, highlighting MicroPython, OpenMV, CV_LITE, and AI Vision (KPU) support.

It integrates the OpenMV image processing framework with MicroPython for efficient programming and leverages the KPU's hardware-accelerated AI vision engine. This makes it a lightweight edge intelligent computing platform for rapid development of AI vision applications, educational experiments, and project verification.

2. What's in the Box

The standard package for the Yahboom K230 AI Development Board includes the following components:

• K230 module
• Screwdriver
• Connection cable
• TF card

Image: All components included in the Yahboom K230 AI Development Board package, including the module, cables, screwdriver, and TF card.

Depending on the specific variant purchased, additional accessories such as fixed brackets, adjustable brackets, or 2DOF electric PTZ components may be included. Refer to your specific product packaging for a complete list.

Image: A detailed table showing the shipping list for the separate K230 module and various optional bracket and PTZ accessories.

3. Product Details and Features

The K230 module is equipped with a range of features to support diverse AI applications:

• High-Performance Chip: 1.6GHz main frequency chip with 6TOPS equivalent computing power.
• Display: 2.4-inch LCD capacitive touch screen with 640x480 resolution.
• Camera: 2MP HD camera for visual input.
• Connectivity: Onboard WiFi module, serial communication interface, and a new 12Pin GPIO interface for expansion.
• Audio: Built-in microphone for audio input and recording, with support for external speakers.
• Memory: 1GB LPDDR4 RAM.
• Storage: Supports TF card expansion for storing models and programs.
• Cooling: Optional cooling fan to maintain stable operation.
• Housing: Resistant ABS case.

Image: An infographic detailing the K230's high computing chip, WiFi, microphone, GPIO, multi-platform support, HD camera, speaker expansion, cooling fan, TF card, capacitive touch screen, and bracket options.

Physical Components

Image: Rear view of the K230 module, showing various ports and connectors.

Image: Close-up view of the K230 module's integrated camera and surrounding components.

Image: Close-up view of the K230 module's USB Type-C port and adjacent pins.

Image: The K230 module mounted on a robot platform, demonstrating its application in mobile robotics.

4. Setup

Before operating the K230 AI Development Board, ensure all necessary components are connected correctly.

4.1 Hardware Connection

1. Power Supply: Connect the K230 module to a 5V power source via the USB Type-C port.
2. TF Card: Insert the provided TF card into the TF card slot on the module. This card contains the pre-installed GUI program and other necessary files.
3. Expansion (Optional): If using external sensors or modules, connect them to the 12Pin GPIO interface as per your project requirements. Refer to the Hardware Distribution Diagram for pin assignments.
4. Bracket Assembly (Optional): If your package includes a bracket, assemble it according to the specific bracket instructions to mount the K230 module securely.

Image: A diagram illustrating the hardware layout of the K230 module, including button locations, ports, and a detailed table of the 12Pin GPIO functions and pin numbers.

4.2 Initial Power-Up

Once all connections are secure, power on the device. The 2.4-inch LCD screen should light up, displaying the pre-installed GUI program.

5. Operating Instructions

The K230 module comes with a pre-installed Graphical User Interface (GUI) program, offering over 30 functional applications. These applications can be accessed directly without additional coding.

5.1 Navigating the GUI

The 2.4-inch capacitive touch screen allows for intuitive navigation. Tap on icons to select applications or adjust settings. The GUI includes functions such as:

• Settings
• AI Face (Face Recognition, Gaze Detection, etc.)
• AI Body (Human Key Point Detection, Fall Detection, etc.)
• AI Scene Applications
• LLM Large Model (Large Language Model interface)
• Camera
• Code Recognition (QR Code, Barcode)
• Color Recognition
• File Manager
• Gallery
• Graphics Detection
• Hardware Interface Test
• Drawing

Image: Screenshot of the K230's pre-installed GUI, showing various application icons and a menu structure for AI functions.

5.2 AI Visual Recognition Applications

The K230 supports a wide array of AI visual recognition tasks. Examples include:

• Color Recognition
• Road Sign Recognition
• Visual Line Patrol
• Face Recognition
• Label Recognition
• QR Code and Barcode Recognition
• Feature Detection
• Digital Recognition
• Object Counting
• License Plate Recognition

Image: A collage demonstrating various AI visual recognition applications such as rectangle detection, object counting, QR code recognition, face key point recognition, face 3D network, human key point detection, fall detection, license plate recognition, and yolov8n target detection.

5.3 Touch Screen Experiments

The capacitive touch screen allows for interactive experiments:

1. Touch Detection: Displays coordinates when the screen is touched.
2. Touch Drawing: Use your finger as a brush to draw on the screen, with options for changing colors and erasing.
3. Touch Photography: Capture images using the HD camera and save them.

Image: Examples of touch screen experiments on the K230, including touch detection showing coordinates, touch drawing with color options, and touch photography capturing an image.

5.4 Real-time Image Transmission

The K230 supports real-time image and video transmission via USB data cable or WiFi, facilitating visual interaction and debugging with a PC.

6. Development Environment

The K230 AI Development Board supports multiple development platforms and programming methods.

6.1 MicroPython Programming

MicroPython is a streamlined implementation of Python 3, optimized for microcontrollers. It offers interactive prompts, integer precision, closures, list derivations, generators, and exception handling. It is designed for compatibility with normal Python, easing code migration from desktop to embedded systems.

6.2 CanMV Development Environment

The K230 visual recognition module utilizes CanMV IDE as its development environment. This IDE provides a simple and intuitive interface for function input parameters and related content. It supports free compilation, online code modification and debugging, and direct download to the K230 module for operation.

Image: Screenshots illustrating the MicroPython programming interface and the CanMV development environment.

6.3 Multi-Controller Compatibility

The K230 can be seamlessly connected to various controllers via its serial communication interface. It supports integration with PC computers, MSPM0, STM32, ESP32, PICO, Raspberry Pi, UNO, Microbit, Jetson, RDK, and other mainstream development boards. This allows for easy output of visual recognition results to external controllers without complex visual algorithm development.

Image: A collage of various development boards such as ESP32, Raspberry Pi, Jetson, and Micro:bit, demonstrating the K230's multi-controller compatibility.

6.4 Model Training and Deployment

Yahboom provides custom model training tutorials and self-developed upper computer toolkits. This includes an online platform for model training and local deployment, offering a complete framework for creating and deploying custom AI models. Users can collect image data, label objects, train models, and then run the code on the K230 to identify targets.

Image: A flowchart illustrating the steps for local model training and deployment, from data collection to target identification.

6.5 Large Language Model Integration

The K230 features a built-in encapsulated large model interaction library, supporting integration with online cloud platforms like OpenRouter. Users can configure API keys to call multiple large language models such as Gemini 2.0 Flash, Deepseek, Llama 4, and Claude 3.7 for advanced interactive functions. Note that some models may require payment for usage.

Image: A visual representation of the K230's large language model integration interface, showing compatibility with various LLM platforms and an example conversation.

7. Specifications

Image: A table summarizing the key product parameters and specifications of the K230 visual recognition module.

Product Size Diagrams

Image: Technical drawings showing the dimensions of the K230 visual recognition module, adjustable angle bracket, 2DOF electric PTZ, fixed bracket, and heightened angle adjustable bracket.

8. Maintenance

To ensure the longevity and optimal performance of your Yahboom K230 AI Development Board, follow these maintenance guidelines:

• Keep Clean: Regularly clean the module with a soft, dry cloth. Avoid using liquids or abrasive cleaners.
• Environmental Conditions: Operate and store the module in a dry environment, away from extreme temperatures, humidity, and direct sunlight.
• Handle with Care: Avoid dropping or subjecting the module to physical shock.
• Firmware Updates: Periodically check the official Yahboom website for firmware updates to ensure you have the latest features and bug fixes.
• Cooling: If using the optional cooling fan, ensure it is free from dust and obstructions to maintain proper airflow.

9. Troubleshooting

If you encounter issues with your K230 AI Development Board, consider the following common troubleshooting steps:

• No Power: Ensure the USB Type-C cable is securely connected to a 5V power source and the module. Try a different cable or power adapter.
• Screen Not Displaying: Verify the module is powered on. If the screen remains blank, try restarting the module.
• Application Errors: If a specific application is not working, ensure the TF card is properly inserted and contains the correct program files. Refer to the documentation for the specific application.
• Connectivity Issues (WiFi/Serial): Check cable connections for serial communication. For WiFi, ensure the module is within range of a network and credentials are correctly entered in the settings.
• Unresponsive Touch Screen: Restart the module. If the issue persists, ensure the screen is clean and free from debris.
• Performance Issues: Ensure the module is adequately cooled, especially during high-load tasks. Close any unnecessary background applications if developing custom code.

For more detailed troubleshooting or persistent issues, refer to the official Yahboom technical support resources.

10. Support and Resources

Yahboom provides extensive resources to assist developers and users:

• Technical Q&A: Access technical support for any questions or issues.
• Lifetime Firmware Updates: Benefit from ongoing firmware updates to enhance functionality and performance.
• Original Model Training Tutorials: Detailed guides for training custom AI models.
• Self-Developed Upper Computer Toolkits: Software tools to aid in development and learning.
• Course Materials: Comprehensive course materials covering various aspects of the K230 module and AI vision.
• Video Tutorials: A library of video tutorials with English subtitles explaining principles, code, demonstrations, and experimental phenomena.

Image: A visual representation of the available video tutorials with English subtitles, covering various topics related to the K230 module.

Image: An overview of the K230 course materials, including topics like CV-lite image processing, face recognition, scenario applications, and detailed code comments.

For further information and to access these resources, please visit the official Yahboom website or contact their technical support team.