---

ESP32-S3 Face Recognition with SD Card Execution

A comprehensive guide to building a self-contained face recognition system using the ESP32-S3, MicroPython, an OLED display, and SD card execution. Btw dont forget to star this <3

---

Table of Contents

1. Introduction
2. Hardware Requirements
3. Software Requirements
4. File Structure
5. MicroPython Installation
6. Steps to Add OV2640 to ESP32
7. How to Run

---

Introduction

This project implements a face recognition system using ESP32-S3. It dynamically loads Python scripts and executes them directly from an SD card, enabling easy updates without re-flashing.

Features:

• Face recognition with an ESP32-S3 camera module.
• Modular components (OLED, SD card, GPIO buttons).
• Supports four modes:
  • On/Off: Long press to toggle power.
  • Register: Capture and register a new face.
  • Cancel: Exit registration mode.
  • Clear: Delete all registered faces.

---

Hardware Requirements

Component	Quantity	Description
ESP32-S3 Board	1	With built-in or external camera.
Micro SD Card + Adapter	1	To store Python files.
SD Card Module	1	For SD card interfacing.
OLED Display	1	I2C-based, SSD1306 128x64.
Buttons	4	For On/Off, Register, Cancel, Clear.
LEDs	2	Green and Red LEDs.
Buzzer	1	For audio feedback.
Resistors	As needed	For button pull-up/down circuits.
Jumper Wires	Several	For connections.
Power Supply	1	ESP32-compatible (e.g., 5V via USB).
OV2640 Cam Module	1	Camera module used for this project.

---

Software Requirements

1. Firmware

• MicroPython for ESP32-S3.

2. Python Tools

Install these tools on your computer:

pip install esptool adafruit-ampy mpremote rshell micropython-uasyncio

3. Libraries for Components

Place these files in the /sd/components/ directory:

• ssd1306.py: OLED Display Driver.
• esp32_camera.py: Camera handling for face recognition (use custom camera build).
• hardware.py: Handles GPIO, buttons, LEDs, and buzzer.

---

File Structure

Organize the SD card as follows:

/sd/
  ├── main.py             # Main program
  ├── components/         # Component drivers and utilities
  │   ├── ssd1306.py      # OLED driver
  │   ├── esp32_camera.py # Camera logic
  │   ├── hardware.py     # Buttons, LEDs, and buzzer
  └── data/               # Registered face data
      ├── encodings/

---

MicroPython Installation

Step 1: Download MicroPython

Download the appropriate MicroPython firmware for ESP32-S3 from here.

Step 2: Flash the Firmware

1. Install esptool:

   pip install esptool

2. Erase existing firmware:

   esptool.py --chip esp32 erase_flash

3. Flash MicroPython:

   esptool.py --chip esp32 write_flash -z 0x1000 micropython.bin

---

Steps to Add OV2640 to ESP32

1. Hardware Connections

The OV2640 camera module has several pins that need to be connected to the ESP32. Here's a common pin mapping:

OV2640 Pin	ESP32 Pin (Example)	Description
VCC	3.3V	Power supply (3.3V).
GND	GND	Ground.
D0 - D7	GPIO32-GPIO39	Data lines.
XCLK	GPIO0	Clock signal.
PCLK	GPIO21	Pixel clock.
VSYNC	GPIO25	Vertical sync.
HREF	GPIO23	Horizontal sync.
SCL	GPIO22	I2C clock.
SDA	GPIO21	I2C data.
RESET	GPIO15 (Optional)	Reset pin.
PWDN	GPIO26 (Optional)	Power down (optional).

• Use jumper wires or a dedicated ESP32 camera development board (e.g., ESP32-CAM) for connection.
• Double-check pin assignments for your specific ESP32-S3 development board, as they may vary.

---

2. MicroPython Firmware

Ensure you are using a MicroPython firmware version that supports the camera interface. Here’s how to get started:

Download Camera-Compatible Firmware

• Use a MicroPython build with camera support, such as:
  • MicroPython Camera Firmware by Loboris.
  • Official ESP32-CAM firmware (check documentation for camera support).

Flash Firmware

Follow the same process as flashing standard MicroPython firmware:

esptool.py --chip esp32 write_flash -z 0x1000 camera_firmware.bin

---

4. Troubleshooting

• Black Image or No Output:

  • Ensure framesize and quality settings match your camera module's capability.
  • Verify pin connections and mappings.
  • Check the power supply (some cameras require a stable 3.3V).

• Camera Initialization Fails:

  • Double-check the firmware compatibility with OV2640.
  • Try reducing the xclk_freq_hz value (e.g., 10 MHz).

• Error: camera module not found:

  • Ensure you have flashed a camera-compatible MicroPython firmware.

---

5. Alternative: ESP-IDF for Advanced Control

For more advanced control or performance, consider using the ESP-IDF framework with native ESP32-CAM libraries, as it provides full support for OV2640.

---

How to Run

1. Prepare the SD Card:

   • Copy main.py and the components/ folder to the SD card.

2. Insert SD Card:

   • Insert the SD card into the SD card module.

3. Power On ESP32:

   • Long press the On/Off button.

4. Modes:

   • Normal Mode: Detects faces.
   • Register Mode: Add a new face.
   • Clear Mode: Deletes all faces.

5. Debug:

   • Use a serial terminal to monitor output.

---

flowchart TD;
    Start["Start / Power On"];
    Start --> HardwareCheck["Check Components"];
    HardwareCheck -->|Hardware OK| MainMode;
    HardwareCheck -->|Hardware Error| ErrorDisplay["Show 'Hardware Error' on OLED"];
    MainMode["Main Mode: Detect Face"] -->|Face Detected| AccessGranted["Green LED Blinks + 'Access Granted' on OLED"];
    MainMode -->|Face Not Detected| AccessDenied["Red LED Blinks + 'Access Denied' on OLED"];
    MainMode --> Register["Register Button Pressed"];
    Register --> CaptureFace["Capture Face"];
    CaptureFace --> ShowRegister["Show 'Register?' on OLED"];
    ShowRegister -->|Cancel Pressed| MainMode;
    ShowRegister -->|On/Off Pressed 3x| SaveFace["Save Face to SD"] --> MainMode;
    MainMode --> ClearMode["Clear Button Pressed"];
    ClearMode --> DeleteFaces["Delete All Faces on SD"] --> MainMode;
    MainMode --> ResetMode["Reset Button Pressed"];
    ResetMode --> Reinitialize["Reinitialize Hardware"] --> MainMode;
    MainMode --> PowerOff["Long Press On/Off"];
    PowerOff --> ShutDown["Show 'Shutting Down' on OLED"];

Loading