Overview

The M-Vision Cam is a programmable smart camera that integrates an image recognition sensor with a microcontroller. It is capable of independently performing basic edge vision processing tasks with low power consumption. Use OpenMV IDE and Micro Python to program the M-Vision Cam for functions such as color tracking, object detection, April tag recognition, and simple image classification.

The M-Vision Cam supports UART communication, allowing it to easily exchange data with host controllers such as the MATRIX Mini R4 and send image recognition results in real time. This enables the Mini R4 to execute control instructions corresponding to image data, such as driving to track targets or automated decision-making. Due to the M-Vision Cam’s built-in computing

power, it can be used an entry-level implementation of AI edge computing and machine vision applications, and is widely used in education, maker projects, and robotics competitions.

Physical Layout and Key Components

Main Components:

1. Camera Lens: Located on the front of the module, captures images

2. Reset Button: Press to restart the Micro Python program running on the camera.

3. RGB LED: Programmable multi-color LED to display operating status or other user-defined functionality.

4. Fill Light LEDs: LEDs on either side of the lens that provide light when ambient lighting is low.

5. DFU (Device Firmware Upgrade Button): Use with the USB-C port to put the module in firmware update mode.

6. USB-C Port: Use to connect to a computer for programming, image transmission, firmware updates, or to power the module.

7. UART Port: Provides serial communication capability for data exchange with the MATRIX Mini R4 or other microcontrollers.

Features

The M-Vision Cam is a smart lens module with built-in image processing capabilities. It can complete image analysis and send results without an external computer. Here are its main features:

⚫ Image processing:

M-Vision can perform a variety of image recognition tasks, such as:

◼ Color block detection and tracking: Find and track specific colors

◼ Line detection: Identify and track lines or edges on the ground

◼ Shape recognition: Detect common shapes like circles and rectangles

◼ April Tag recognition: Accurately read visual tag ID

◼ Simple face detection

◼ QR Code scanning

⚫ Programming development environment:

◼ Micro Python: Use concise and easy-to-learn Python syntax to program the camera

◼ OpenMV IDE: Write programs, preview camera images, and adjust parameters in real time on a computer

◼ Main program execution: Name the main program main.py and load it onto the camera. It will automatically execute after booting.

⚫ Data transmission and communication:

◼ UART serial communication: Transmits identification results (such as coordinates or tag ID) to the main control board

◼ USB Virtual Serial Port (VCP): When connected to a computer, data can be transmitted or debugged via USB

⚫ Adjustable image settings:

Provides flexible image capture settings to meet the needs of various applications and

scenarios.

◼ Resolution switching: Supports 320x240 (QVGA) and 160x120 (QQVGA)

◼ Pixel format switching: Can select color (RGB565) or grayscale image

◼ Screen flip: Supports horizonal and vertical mirroring (use set_vflip() and

set_hmirror())

◼ Manual parameter settings: Exposure, gain, white balance, etc. (depending on sensor and firmware support)

⚫ Onboard auxiliary components:

◼ RGB status light: Customizable color and flashing, displays program status/debugging

◼ LED fill light: Provides lighting in dark settings, adjust brightness automatically with fill_light.py

◼ Reset button: Restarts the main program

◼ DFU button: Enter firmware update mode

Application Scenarios

⚫ Basic applications: Automatic tracking and tracing

◼ Color tracking car: Detects position of user-designated color blocks and automatically moves the car toward the block.

◼ Line-following robot: Identify black lines and drive car along them.

◼ Obstacle avoidance: Use color block or edge detection to determine locations of obstacles.

⚫ Advanced applications: identification, interaction, and classification

◼ Automatic color sorting: Identify objects of different colors and push them to different areas.

◼ April Tag precise positioning: Use visual tags to assist the robot in navigating, entering a station, or completing a task.

◼ QR Code command start: Read the QR code to perform a specific task, like playing a sound or moving a mechanism.

◼ Face detection interaction: Activate motor, voice, or screen prompts after sensing a face.

⚫ Special application: Interactive system and AI edge computing

◼ Automatic track and shoot: After the camera detects the target, it controls the servo mechanism to track and shoot the target.

◼ Interactive pet device: Tracks a designated object (such as a toy ball) and automatically activates a feeder or makes a sound.

◼ AI model interface (advanced): M-Vision Cam supports TensorFlow Lite for

Microcontrollers. Experienced users can deploy self-trained models for image

classification or recognition tasks.

Electrical Characteristics & Pinout

⚫ M-Vision Cam Hardware S pecifications:

Glossary:

1MByte = 8Mbit。

Image sensor* Different sensor models may vary in low-light performance, dynamic range, shutter type (global/rolling), etc.

⚫ Pin and wiring description