Skip to content

IMX500 MCU AI Camera

Mode Introduction

Arducam MCU AI Camera Module (B0564) supports SPI / USB interfaces and two operating modes:

Deploy Mode

Main Features:

  • No image output
  • Works with both SPI and USB interfaces
  • Outputs only AI inference data
  • Suitable for model deployment

Preview Mode

Main Features:

  • USB interface only
  • Outputs image data
  • Preview camera feed with IMX500PreviewGUI
  • Suitable for installation, debugging, and image preview

Quick Start Guide

Get AI inference data via USB

  • Hardware Preparation

    1. Arducam MCU AI Camera Module (SKU: B0564)
    2. USB Type-C Cable
    3. Windows PC Device

  • Install Dependencies

cd arducam_imx500_mcu_serial_example
pip install -r requirements.txt
  • Run the example
python serial_deploy_mode_demo.py

After starting the program, please use the arrow keys to select the serial port and press Enter.

arducam_wiki_imx500_mcu_ai_camera_qsg_picture_1

You will then see bounding boxes drawn based on AI inference results:

arducam_wiki_imx500_mcu_ai_camera_qsg_picture_2

Get AI inference data via SPI

  • Hardware Preparation

Note

1. The following demonstration is based on Arduino Uno R3 Development Board.  
2. The multi-camera adapter board is designed for testing multiple B0564 camera modules simultaneously. When using a single module, you can connect it directly to the main board using an SPI cable.
  1. Arducam MCU AI Camera Module (SKU: B0564)
  2. Arduino Uno R3 board
  3. Arducam Multi-Camera SPI Adapter Board for Arduino Uno R3
  4. 6-pin SPI Cable
  5. USB Type-B Cable
  6. Windows PC Device

This example uses Arduino Uno R3 to receive AI data via SPI and forward it to the USB serial port.

arducam_wiki_imx500_mcu_ai_camera_qsg_picture_8

  • Hardware Connection

Connect Arduino Uno R3 to the PC using a Type-B USB cable.

  • Select Arduino IDE

    1. Please install Arduino IDE Tool
    2. Open the project arducam_imx500_mcu_spi_arduino_uno_r3_example/main/main.ino in Arduino IDE.
    3. Select the correct board in Arduino IDE and upload the firmware.
      arducam_wiki_imx500_mcu_ai_camera_qsg_picture_3
    4. Wire the devices as shown below:
      Connect Arducam MCU AI Camera (B0564) to Arduino Uno R3 Board:
      arducam_wiki_imx500_mcu_ai_camera_qsg_picture_9

Once wired all the devices, please refer to the Get AI inference data via USB to view inference results on the PC.

Image Preview Mode

There is a **Mode** button on the back of the Arducam IMX500 MCU AI Camera Module.

  • Hold down the **Mode button** while connecting the camera to the PC with a Type-C USB cable
  • The red LED on the side will turn on, indicating Preview Mode

⚠️Some older versions of Windows OS may require libusb driver installation. If so, please download zadig, run it, and install as shown below:

arducam_wiki_imx500_mcu_ai_camera_qsg_picture_4

Download preview tool: Go to IMX500 SPI Examples - Releases to get IMX500PreviewGUI.exe.

After launching the software:

  1. Click the Open button
  2. Click the Play button to preview the camera in real time

arducam_wiki_imx500_mcu_ai_camera_qsg_picture_5

arducam_wiki_imx500_mcu_ai_camera_qsg_picture_10

Model Switching

  1. Connect the camera to the PC via Type-C USB. A flashing green LED indicates Deploy Mode.
  2. In Deploy Mode, long press the Mode button (>2s), and the PC will mount a USB drive named IMX500.
  3. Copy a .adm model file to the drive to complete model switching.

arducam_wiki_imx500_mcu_ai_camera_qsg_picture_6

.adm model files are available in the arducam_imx500_model_zoo project.

For example you can switch the yolov8n_pos adm model to implement human pose estimation applications.

arducam_wiki_imx500_mcu_ai_camera_qsg_picture_7

More Resources

Arducam AI Model Zoo

Github - Arducam AI Model Zoo Source Code

Arducam AI Model Converter Service

Partnership with SONY - Arducam IMX500 AI Camera Solution