project of face detection Python v.1

import cv2
import mediapipe as mp

# Initialize MediaPipe Face Mesh
mp_face_mesh = mp.solutions.face_mesh
face_mesh = mp_face_mesh.FaceMesh()

# Capture video from the webcam
cap = cv2.VideoCapture(0)

while True:
ret, frame = cap.read()
if not ret:
break

# Convert frame to RGB (MediaPipe requires RGB images)
rgb_frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)

# Process the frame to detect facial landmarks
results = face_mesh.process(rgb_frame)

if results.multi_face_landmarks:
for face_landmarks in results.multi_face_landmarks:
for landmark in face_landmarks.landmark:
# Get the x and y coordinates of each landmark
x = int(landmark.x * frame.shape[1]) # Scale to frame width
y = int(landmark.y * frame.shape[0]) # Scale to frame height
# Draw each landmark as a small green dot
cv2.circle(frame, (x, y), 1, (0, 255, 0), -1)

# Display the frame with detected landmarks
cv2.imshow(“Face Landmarks”, frame)

# Exit the loop if ‘q’ is pressed
if cv2.waitKey(1) & 0xFF == ord(‘q’):
break

# Release resources
cap.release()
cv2.destroyAllWindows()

 

Now explanation, this code detects faces in real-time using a webcam and identifies 68 facial landmarks (key points) on each detected face. It uses OpenCV for video capture and image processing, and Dlib for face detection and landmark prediction.

Step-by-Step Description

  1. Import Libraries:
    • cv2: OpenCV for video capture and image processing.
    • dlib: For face detection and facial landmark prediction.
  2. Load Models:
    • detector = dlib.get_frontal_face_detector(): Initializes Dlib’s face detector.
    • predictor = dlib.shape_predictor("shape_predictor_68_face_landmarks.dat"): Loads the pre-trained model for detecting 68 facial landmarks.
  3. Capture Video:
    • cap = cv2.VideoCapture(0): Starts capturing video from the default webcam.
  4. Main Loop:
    • The loop continuously reads frames from the webcam.
    • Converts each frame to grayscale (gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)) to improve face detection performance.
    • Detects faces in the grayscale frame using Dlib’s face detector (faces = detector(gray)).
  5. Detect Facial Landmarks:
    • For each detected face, the code predicts 68 facial landmarks using Dlib’s landmark predictor (landmarks = predictor(gray, face)).
    • Each landmark is drawn as a small green circle on the frame using cv2.circle.
  6. Display Results:
    • The frame with detected faces and landmarks is displayed in a window (cv2.imshow("Face Landmarks", frame)).
  7. Exit Condition:
    • The loop continues until the user presses the q key, at which point the program exits.
  8. Release Resources:
    • The webcam is released (cap.release()) and all OpenCV windows are closed (cv2.destroyAllWindows()).