K230 2MP AI Vision Module CanMV Face Recognition Camera Sensor Module Standard Version Acrylic Shell

K230 2MP AI Vision Module CanMV Face Recognition Camera Sensor Module Standard Version Acrylic Shell

Advantages:
* K230 Vision Recognition Module
* Performance up to 13.7 Times that of K210
* 6 TOPS Equivalent Computing Power
* 1GB RAM
* AI Vision Recognition
* Ideal Tool for Electronics Competitions
* Online Large Language Model Support
* Local Deployment for Model Training
* 108 Code-Level Video Tutorials

Product Features:
Next-Generation K230 AI Computing Platform
* A RISC-V architecture-based edge AIoT chip
* 6 TOPS equivalent computing power
* Third-generation KPU: high precision, fast startup, and ultra-low power consumption
* Developed on the RT-Smart real-time operating system

Exclusive Tool for Electronics Competitions
* Serial communication, compatible with multiple controllers: MSPM0, ESP32, STM32 and Raspberry Pi
* Developed with MicroPython
* Supports various mounting brackets and 2-axis rotating gimbals

1GB LPDDR4 – Ultra-Large Memory
* Easily handles AI vision applications

AI Vision Application Development
* On-device AI vision and AIoT training + local deployment
* Online toolkits support development and validation

108 Code-level Video Tutorials
* Yahboom has meticulously created 108 code-level video tutorials. The videos include principle explanations, code walkthroughs, hands-on demonstrations, and experimental phenomena, helping developers easily master the full-stack development of the K230 vision module.
* Whether you're a beginner or an engineer, you'll find the right course content

Integrated Interface for Large Language Model APIs
* This product comes with a highly encapsulated large model interaction library. Based on online cloud platforms such as OpenRouter and iFLYTEK Spark, users only need to configure an API Key and specify the model type to seamlessly call multiple mainstream large language models (for Gemini 2.0 Flash, DeepSeek V3, LLama 4, and Claude 3.7), achieving efficient interactive functions.

Model Training and Local Deployment
* Yahboom provides step-by-step tutorials for custom model training, including online platform model training and local deployment training environments. It offers a complete process framework, detailed tutorial guidance, and technical support services to ensure every user can easily train their own models. Through the model library and open platform for custom model training, users can train AI models as needed without the need for model conversion steps, directly obtaining models in the kmodel format. In addition, the platform offers example code to help users quickly complete model deployment.

Provides Host Computer Learning Tools
* It includes Yahboom's self-developed host computer learning tool, integrating K230 development and learning resources. It supports RTSP real-time video streaming monitoring and incorporates modules such as video/image processing tools, data generators, and text tools, offering a one-stop development and learning experience.

Supports Audio Processing
* The K230 integrates a microphone and supports audio recording, enabling keyword wake-up and speech-to-text functions. Combined with large language models, it can engage in voice dialogue, providing an intelligent interactive experience.

AIoT Network Applications
* The K230 vision module supports various AIoT application scenarios, providing wireless network connectivity, AP hotspot mode, and multiple network communication protocols. It supports RTSP real-time video transmission and can combine AI image processing to achieve intelligent transmission. It offers intelligent voice/text interaction capabilities to meet diverse IoT development needs.
* Supports Wi-Fi/AP Hotspot Mode
* Supports Multiple Network Communication Protocols
* RTSP Wireless Video Transmission
* Speech-to-Text / Text-to-Speech

Supports Multiple Development Platforms
* The module supports connection to mainstream controllers such as PC, MSPM0, STM32, ESP32, PICO, Raspberry Pi, Jetson, RDK, and others for development

Using MicroPython Programming
* MicroPython is a streamlined and efficient implementation of the Python 3 programming language, optimized for microcontrollers and constrained environments. It provides advanced features such as interactive prompts, arbitrary precision integers, closures, list comprehensions, generators, and exception handling, while being compact enough to run within only 256KB of code space and 16KB of RAM. MicroPython aims to be as compatible as possible with standard Python to facilitate migrating code from desktop environments to microcontrollers or embedded systems.

CanMV Development Environment
* The K230 vision recognition module uses CanMVIDE as its development environment, which is simple and convenient to operate. The IDE provides function input parameter hints and related information, with no need for code compilation. Users can modify and debug code online and directly download the code to the K230 module for execution.

After-Sales Technical Support
* Comprehensive after-sales service to help you solve all kinds of technical challenges

Hardware Features:
* K230 High-Performance Chip: It utilizes the K230 dual-core RISC-V high-performance processor with a 6 TOPS equivalent computing power KPU. Its overall performance is approximately 13.7 times that of the K210.
* High-Definition, High-Frame-Rate Camera: It uses a 1080P 60FPS high-definition digital-level camera. Compared to traditional OV5640/OV5647 sensors, it produces clearer images and better recognition results.
* 1GB Large Memory: With 1GB LPDDR4, it is suitable for various AI vision applications and scenarios requiring large memory.
* Onboard WiFi Module: It connects to networks via built-in WiFi, supporting AIoT applications and real-time RTSP camera streaming.
* Onboard Microphone: The K230 includes an onboard microphone that collects voice input, enabling large language model dialogues and providing intelligent interactive capabilities.
* GPIO Expansion Interface: It features an external 12-pin GPIO interface. Users can customize pin functions to work with various electronic modules.
* Multi-Platform Communication Connectivity: Through extended serial ports, visual recognition results can be sent to host computers, MSPM0, STM32, ESP32, PICO, Raspberry Pi, and other main controllers for efficient interaction.

Application Cases:
* Autonomous Car Driving
* Road Sign Recognition and Turning
* License Plate Recognition
* QR Code Controlled Car
* Face Recognition/Tracking
* Red/Green Light Spot Tracking
* Intelligent Garbage Sorting
* Object Tracking
* Tag Code Following

Specifications:
* Model: K230 Vision Recognition Module
* Version: Standard or Deluxe Version
* CPU 0: RISC-V Processor, 0.8GHz, 32KB I-cache, 32KB D-cache, 128KB L2 Cache
* CPU 1: RISC-V Processor, 1.6GHz, 32KB I-cache, 32KB D-cache, 256KB L2 Cache, 128-bit RVV 1.0 Extension
* KPU: 6TOPS and it supports INT8 & INT16. Typical Network Performance: ResNet 50: ≥ 85fps at INT8; MobileNet_v2: ≥ 670fps at INT8; YoloV5S: ≥ 38fps at INT8
* DPU: 3D structured light depth engine, supports up to 1920×1080 resolution
* Development Environment: MicroPython programming, CanMV IDE
* RAM: 1GB LPDDR4
* VPU: H.264 & H.265 encoder/decoder, max resolution 4096×4096; Encoder performance: 3840×2160 at 20fps; Decoder performance: 3840×2160 at 40fps; JPEG codec: supports up to 8K (8192×8192) resolution
* Onboard Interfaces: Speaker interface, fan interface, Type-C port, serial communication port, GPIO expansion port, and TF card slot
* Image Input: Built-in 2MP HD camera (GC2093)
* Onboard Features: Microphone, buzzer, RGB light and WiFi card
* Power Supply: 5V input via Type-C/communication port/GPIO
* Operating Current: Standard Version: 0.6A; Deluxe Version: 0.8A
* Operating Power: Standard Version: 3W; Deluxe Version: 4W
* Buttons: Reset button + custom function button
* Dimensions: Standard Version: 75 × 47.63 × 19.72mm; Deluxe Version: 72 × 46.37 × 18.9mm
* Weight: Standard Version: 48.9g; Deluxe Version: 64.9g

Packing List:
* 1 x K230 Vision Module
* 1 x Type-C Data Cable (1m)
* 1 x Screwdriver
* 1 x PH2.0 4-Pin Flat Cable – Dual-Head – All Black – Reversed (300mm)
* 1 x Bundle of PH2.0 4-Pin to Dupont Loose Wires (Male to Female, 200mm)
* 1 x User Manual
* 1 x After-Sales Service Card
    
Packaging Details:
* Weight: 0.2kg

Note:
* 32GB TF card and bracket are not included. Please prepare your own TF card to flash the GUI program.