Face Detection and Tracking Project Based on Dlib

This is a repository based-dilb for showing simple examples of face recognition based on the Dlib library

---

​ This project contains several Python scripts focused on face detection, landmark detection, face recognition, and tracking using the dlib and cv2 libraries. Below is an overview of each script and its primary functionality.

⭐ Star the repository on GitHub — it motivates me a lot!

Scripts Overview

1. Face Detection

• face_detection/dlib_frontal_face_detector_img.py:
  • Detects faces in a single image using dlib's frontal face detector.
  • Displays and saves the result with detected faces outlined with rectangles.
• face_detection/dlib_frontal_face_detector_video.py:
  • Detects faces in a real-time video stream from a webcam using dlib's frontal face detector.
  • Displays the detected faces with bounding rectangles and shows the current frames per second (FPS).
• Haar_Cascade/CascadeClassifier_img.py:
  • Uses OpenCV's Haar Cascade Classifier to detect faces in an image.
  • Displays the image with detected faces outlined.

2. Face Landmarks Detection

• face_landmark_detection_img.py:
  • Detects facial landmarks (e.g., eyes, nose, mouth) on faces in an image using dlib's pre-trained models.
  • Displays the image with detected landmarks marked.
• face_landmark_detection_video.py:
  • Detects facial landmarks on faces in a real-time video stream using dlib's pre-trained models.
  • Displays the video feed with detected landmarks and FPS information.
• face_recognition.py:
  • Recognizes faces in an image by detecting facial landmarks.
  • Displays the image with recognized landmarks highlighted.

3. Face and Object Tracking

• face_tracking_dlib.py:
  • Tracks a face in real-time using dlib's correlation tracker and a webcam.
  • Displays the video feed with tracking rectangles and FPS information.
• face_tracking_info.py:
  • Similar to face_tracking_dlib.py, but includes additional on-screen information about tracking status.

face_tracking.mp4

• object_tracking_dlib.py:
  • Provides a user-interactive tracking system where the user selects an area in a video feed to be tracked.
  • Supports starting and stopping the tracking process with keyboard inputs.

Firstly, Use the mouse to draw an object detection box
Then:
'1': starting tracking, '2': stop tracking, 'Esc': exit 

---

Getting Started

Prerequisites

• Python 3
• Install required packages:

dlib==19.24.2
face-recognition==1.3.0
face-recognition-models==0.3.0
opencv-python==4.8.0

Models

This project utilizes several pre-trained model files, including Haar feature classifiers and Dlib's facial landmark predictors. These model files have been uploaded to Google Drive and Baidu Cloud(extraction code: dlib) for users to download as needed.

Instructions

Download and extract the model files into the models folder within the local project directory.

models
 ├── haarcascade_eye.xml
 ├── haarcascade_eye_tree_eyeglasses.xml
	······
 ├── shape_predictor_5_face_landmarks.dat
 └── shape_predictor_68_face_landmarks.dat

Running the Scripts

• To run face detection on an image:

  python face_detection/dlib_frontal_face_detector_img.py

• To start real-time face detection using the webcam:

  python face_detection/dlib_frontal_face_detector_video.py

• To use track function ( including face tracking and object tracking):

  python face_tracking/face_tracking_dlib.py	# for face tracking
  python face_tracking/object_tracking_dlib.py	# for object tracking

In the End

If you have any questions, feel free to post them in the issues section or contact me directly.

If this repository is useful to you, I will appreciate you will star this repository.