MU Vision Sensor

  • 999 kr

Open the intelligence era, starting with visual recognition! Here comes an image recognition sensor with a built-in deep learning engine. Small in size, the sen

I lager.
Lagersaldo: 2
Artikelnummer: 515

Open the intelligence era, starting with visual recognition! Here comes an image recognition sensor with a built-in deep learning engine. Small in size, the sensor adopts 30W wide-angle lens and various vision algorithms, which could make your robots more intelligent and independent. With this MU vision sensor, you can build up a smart robot that can trigger suitable behaviors to deal with random events in the environment.

Use this MU vision sensor to recognize and locate a variety of objects, such as colors, balls, humans, and cards. The detected result can be output through UART and IIC. It can process information locally without a network connection. In addition, the parameter configuration and firmware update of the module can be directly realized via the onboard USB serial port. The sensor supports UART, IIC, and WIFI communication modes. What’s more, it is suitable for Arduino, Mixly, and micro: bit embedded platforms, so you are always welcomed to explore the powerful capability of this sensor regardless of the development platforms.

This small MU vision sensor can help you to complete intelligent applications, for instance, smart printing and scanning car robot, basketball shooting robot, and so on. Moreover, it can be widely used in smart toys, AI teaching tools, maker products, etc.

Project: How To Make DIY Arduino Traffic Sign Recognition Robot

Key Features

  • Small Footprint, Fast Processing Speed, and Low Power-Consumption
  • Good Compatibility: It can be supported on many different platforms like Arduino, Mixly, and Micro: bit.
  • Multiple user modes supported: The MU vision sensor supports UART, I2C, WIFI communication mode, and image transmission mode.
  • No network necessary: All the visual recognition algorithms on the MU vision sensors are processed locally, with no need for the network, and no worries about privacy issues.


  • Human Body Detection (upper body)
  • Ball Detection (orange table tennis ball and green tennis ball)
  • Card Recognition (20 pieces: 5 traffic-sign cards, 5 shape cards, 10 number cards)
  • Color Block Detection (outputs length, width, and the center position of the specified color block in the camera)
  • Color Recognition (outputs RGB information of the specified point in the camera)
  • WIFI Data Transmission
  • WIFI Image Transmission
  • Motion Gesture Detection
  • Proximity Detection
  • Ambient Light Detection 


  • Processor: dual-core, 240MHz
  • Camera: Omnivision ov7725
  • Camera Resolution: 640x480
  • Field of View: 90° (Diagonal line)
  • 2 LED lights
  • Dimension: 3.2*3.2*1.2cm/1.26*1.26*0.47”
  • Data Output: UART/IIC
  • Power Supply: 5V
  • Detected Objects: people silhouette (above the waist)

                Ball (Ping-pong ball or tennis ball)                      

                20 Customized Cards                    

               Color (Within the detection range, output the color of the designated position or output the position of the  designated color) 

  • Software Library: milxy’s library functions and reference programs

                Arduino library functions                      

                Micro:bit library functions

  • Functions In Development: WiFi Image Transmission (module as AP)

                 Motion Routine Detection(detection of hand movement from top to bottom, left to right, etc,. non-gesture  sensing)

                 Face recognition                                

                 QR code recognition

ProcessorESP32, Dual-core,240MHz
Camera Resolution85° wide-angle, 640x480
Power Supply Voltage
(by 4-Pin PH2.0 Interface)
Data Output InterfaceUART / I2C / WIFI
Supported IDEsArduino
mPython X
Supported ControllersArduino
Micro: bit


  • MU Vision Sensor x1