Face Recognition Door Lock using ESP32-CAM

Upgrade your smart lock system with face recognition for secure, contactless access.
This project uses ESP32-CAM to detect and recognize faces and controls a servo motor to unlock doors for authorized users.

---

🛠️ Hardware Requirements

• ESP32-CAM Module (with OV2640 camera)
• Servo Motor or Electromagnetic Lock
• FTDI Programmer (for programming ESP32-CAM)
• Jumper wires and breadboard
• 5V power supply for servo
• Optional: OLED display for lock status feedback

---

⚡ Working Principle

1. Face Detection & Recognition:

   • ESP32-CAM captures frames and detects faces using ESP32 face recognition libraries (like ESP-WHO).
   • Recognized faces are matched against a stored database of authorized users.

2. Lock Control:

   • If a face is recognized, ESP32 triggers the servo motor to unlock the door.
   • If unauthorized, the door remains locked and optionally triggers an alert.

3. Optional Wi-Fi Monitoring:

   • Lock/unlock events and unauthorized access attempts can be sent to a cloud dashboard or app.

---

💻 Arduino/ESP-IDF Code Example (ESP-WHO)

Note: Face recognition requires ESP-WHO library in ESP-IDF environment.
Below is a simplified logic outline:

#include "esp_camera.h"
#include "soc/soc.h"
#include "soc/rtc_cntl_reg.h"
#include <Servo.h>

#define SERVO_PIN 23
Servo lockServo;

// Camera pins configuration for ESP32-CAM
camera_config_t config = {
    .pin_pwdn = -1,
    .pin_reset = -1,
    .pin_xclk = 0,
    .pin_sccb_sda = 26,
    .pin_sccb_scl = 27,
    .pin_d7 = 35,
    .pin_d6 = 34,
    .pin_d5 = 39,
    .pin_d4 = 36,
    .pin_d3 = 21,
    .pin_d2 = 19,
    .pin_d1 = 18,
    .pin_d0 = 5,
    .pin_vsync = 25,
    .pin_href = 23,
    .pin_pclk = 22,
    .xclk_freq_hz = 20000000,
    .pixel_format = PIXFORMAT_JPEG,
    .frame_size = FRAMESIZE_QVGA,
    .jpeg_quality = 10,
    .fb_count = 1
};

void setup() {
  Serial.begin(115200);
  lockServo.attach(SERVO_PIN);
  lockServo.write(0); // Lock closed

  // Initialize camera
  esp_err_t err = esp_camera_init(&config);
  if (err != ESP_OK) {
    Serial.printf("Camera init failed: %d\n", err);
    return;
  }

  Serial.println("Face recognition door lock ready!");
}

void loop() {
  // Capture frame
  camera_fb_t *fb = esp_camera_fb_get();
  if(!fb){
    Serial.println("Camera capture failed");
    return;
  }

  // TODO: Use ESP-WHO API to detect face
  bool faceDetected = true; // Replace with actual detection
  bool authorizedFace = true; // Replace with recognition logic

  if(faceDetected){
    if(authorizedFace){
      Serial.println("Authorized face detected! Unlocking...");
      lockServo.write(90); // Unlock
      delay(5000);
      lockServo.write(0); // Lock again
    } else {
      Serial.println("Unauthorized face detected!");
      // Optional: Wi-Fi alert
    }
  }

  esp_camera_fb_return(fb);
  delay(2000);
}

ESP32-CAM Face Recognition: Understanding faceDetected and authorizedFace

In ESP32-CAM face recognition projects, you often see placeholder code like this:

// TODO: Use ESP-WHO API to detect face
bool faceDetected = true; // Replace with actual detection
bool authorizedFace = true; // Replace with recognition logic

Detect a Face

Use ESP-WHO’s face detection API:

face_id_list_t face_list;
mtmn_config_t mtmn_config = mtmn_init_config();
mtmn_config.type = FAST;

// Convert camera frame to RGB format
dl_matrix3du_t *image_matrix = dl_matrix3du_alloc(1, fb->width, fb->height, 3);
fmt2rgb888(fb->buf, fb->len, fb->format, image_matrix->item, fb->width, fb->height);

// Perform detection
face_list = face_detect(image_matrix, &mtmn_config);

if(face_list.length > 0){
    faceDetected = true;
} else {
    faceDetected = false;
}

dl_matrix3du_free(image_matrix);

Recognize the Face

ESP-WHO generates face embeddings, which are numeric representations of faces. Compare the detected face embedding with stored embeddings:

if(faceDetected){
    float similarity = face_recognize(face_embedding, authorized_embedding);
    if(similarity > 0.7){ // Threshold for recognition
        authorizedFace = true;
    } else {
        authorizedFace = false;
    }
}

Connect servo motor to ESP32-CAM pin (e.g., GPIO23).

Use programmer to upload code to ESP32-CAM.

Train ESP32-CAM with authorized face images using ESP-WHO.

Place ESP32-CAM at the door entrance.

When a face is recognized, the servo unlocks for a few seconds, then locks again.