---

🎯 Face Recognition with Python: Using the face_recognition Library by ageitgey

Security systems are getting smarter every day — and face recognition is one of the leading technologies behind this evolution. Thanks to the amazing work of Adam Geitgey, we have a powerful yet simple-to-use Python library called face_recognition, which allows you to build facial recognition systems with just a few lines of code.

In this blog, we’ll walk through how to use the face_recognition library to:

• Load images of known people.
• Identify faces in new images or webcam feeds.
• Match unknown faces against a stored database.

---

📦 What Is face_recognition?

face_recognition is a Python library that uses deep learning models (via dlib) under the hood to identify and compare human faces. It is:

• Open-source
• Accurate
• Beginner-friendly
• Cross-platform (Windows, Linux, macOS, Raspberry Pi)

---

🔧 Installation

Before you begin, you need Python 3 and pip. Then, simply run:pip install face_recognition pip install opencv-python

Note: You may need dlib dependencies. On Raspberry Pi or Linux, install via system packages first (e.g., libdlib-dev).

---

🧠 How Face Recognition Works

The library uses a pre-trained deep learning model to extract 128-dimensional face encodings from images. These encodings represent unique features of a person’s face.

Matching is done by comparing face encodings using Euclidean distance — if it’s small enough, the faces are considered a match.

---

📁 Project Structure

face_recognition_project/ ├── known/ │ ├── obama.jpg │ └── biden.jpg ├── unknown/ │ └── group_photo.jpg └── recognize_faces.py

---

🧪 Python Code Example

import face_recognition import cv2 import os # Load known face encodings known_face_encodings = [] known_face_names = [] for file in os.listdir("known"): image = face_recognition.load_image_file(f"known/{file}") encoding = face_recognition.face_encodings(image)[0] known_face_encodings.append(encoding) known_face_names.append(file.split('.')[0]) # Load an unknown image unknown_image = face_recognition.load_image_file("unknown/group_photo.jpg") face_locations = face_recognition.face_locations(unknown_image) face_encodings = face_recognition.face_encodings(unknown_image, face_locations) # Draw results for (top, right, bottom, left), face_encoding in zip(face_locations, face_encodings): matches = face_recognition.compare_faces(known_face_encodings, face_encoding) name = "Unknown" if True in matches: first_match_index = matches.index(True) name = known_face_names[first_match_index] # Draw a rectangle and label cv2.rectangle(unknown_image, (left, top), (right, bottom), (0, 255, 0), 2) cv2.putText(unknown_image, name, (left, top - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.9, (255, 255, 255), 2) # Convert BGR to RGB and show cv2.imshow("Face Recognition", cv2.cvtColor(unknown_image, cv2.COLOR_RGB2BGR)) cv2.waitKey(0) cv2.destroyAllWindows()

---

🖼️ Example Image (Generated)

Let’s visualize what the system detects when scanning a group photo with known faces.

Here’s a simulated result:

Note: Green boxes show recognized faces, with labels identifying them.

---

💻 Real-World Applications

• Home Security Systems – Raspberry Pi can trigger events when an unknown face is detected.
• Office Entry Systems – Automatically unlock doors for authorized personnel.
• School/College Attendance – Face-based roll calls.
• Retail Surveillance – Track VIP or blacklisted customers.

---

✅ Proof of Use

The library is actively used in Raspberry Pi-based facial recognition projects and is maintained on GitHub:
🔗 https://github.com/ageitgey/face_recognition

---

🚀 Final Thoughts

Whether you’re a hobbyist working on a Raspberry Pi project or an engineer building intelligent surveillance systems, face_recognition gives you the power of deep learning with minimal code.

---