How mstfydz/opencv-anlik-yuz-tanima works
This is a small OpenCV real-time face-recognition application, organized as separate scripts for collecting face samples, training a recognizer, and identifying faces from a live camera. Its main workflow saves detected face crops in veriseti, trains an LBPH model, then loads that model during live recognition and overlays names and confidence. The README confirms camera-based recognition; the included cascade files support face detection. The egitim model directory is referenced by code but is absent from the supplied tree.
Main entry points: 02_yuz_egitimi.py Train Model · 01_yuz_veriseti.py Capture Faces · haarcascade_frontalface_default.xml Face Cascade · veriseti Face Dataset
Components 11
Sample Capture
- Capture Faces
01_yuz_veriseti.py - Face Dataset
veriseti
Model Training
- Train Model
02_yuz_egitimi.py - LBPH Model
Live Recognition
- Recognize Faces
- Text Overlay
Camera Support
- Face Cascade
Cascades/haarcascade_frontalface_default.xml - Eye Cascade
Cascades/haarcascade_eye.xml - Camera Preview
simpleCamTest.py
Other components
- Operator
- Camera
Connections 15
- Operator → Capture Faces: starts capture
01_yuz_veriseti.py - Camera → Capture Faces: provides frames
01_yuz_veriseti.py - Capture Faces → Face Cascade: detects faces
01_yuz_veriseti.py - Capture Faces → Face Dataset: writes crops
01_yuz_veriseti.py - Operator → Train Model: starts training
02_yuz_egitimi.py - Train Model → Face Dataset: reads samples
02_yuz_egitimi.py - Train Model → Face Cascade: detects faces
02_yuz_egitimi.py - Train Model → LBPH Model: writes model
02_yuz_egitimi.py - Operator → Recognize Faces: starts recognition
- Camera → Recognize Faces: provides frames
- Recognize Faces → Face Cascade: detects faces
- Recognize Faces → LBPH Model: reads model
- Recognize Faces → Text Overlay: renders labels
- Operator → Camera Preview: starts preview
simpleCamTest.py - Camera → Camera Preview: provides frames
simpleCamTest.py