Overview

This project implement the functions of matlab equivalent MTCNN+ArcFace face recognition+tracking+HC08 Bluetooth module transmitting/receiving signal+voice broadcast, etc. The code implementation is relatively simple, condensed arduino control program+matlab algorithm design, and has a strong understanding.

Hardware and Software

• usb camera
• 2 servo
• arduino
• arduino support package
• blueTooth HC08
• matlab2020b+

Advantages

• Simple code, integrated control + algorithm.
• It is easy to debug the verification algorithm repeatedly.
• Fast development cycle.

Disadvantages

• Compared with C++, the execution efficiency is lower
• Not easy to transplant
• SOTA algorithm unified integration into matlab is slow

How to use

1. Download arduino support package;
   
2. Configure the hardware environment, such as arduino wiring, steering gear wiring, jetson maximum power supply use short-circuit plug, the default is usb power supply, etc.;
   
3. Prepare face data, interact with source/prepare_getFaces.m on the pc, then execute source/prepare_faceDatabase.m to save database features, and finally run source/camera_faceRec.m
   

bluetooth control

Refer to here for the hardware connection diagram

Reference

Arduino-matlab
jetson_faceTrack_pytorch
Face Detection and Alignment MTCNN

notes

author:cuixingxing
email:cuixingxing150@gmail.com