Automatic Garage Door

Deskripsi

Computer vision project menggunakan OpenCV

Algoritma: Yolo (real-time object detection)

Instalasi

• Python
• Python pip
• Open CV pip install opencv-python
• Tesseract pip install pytesseract and Tesseract OCR Executables (Required)
• imutils pip install imutils
• YoloV7 Weights and config
• Aplikasi DroidCam

Cara menggunakan

• Letakan YoloV7 Weights and config pada folder /YOLO/
• Letakan video pada folder /resource/videos
• Jalankan program menggunakan perintah python main.py pada terminal
• Ubah source feed pada kode

   path = 'resource/videos/lands1.mp4'
   # File: 'Videos/car.mp4' 
   # Network: 'http://192.168.1.51:4747/video'
  
   cap = cv2.VideoCapture(path)
   # cap = cv2.VideoCapture(0) # Webcam

• Apabila menggunakan DroidCam sebagai sumber, ganti linknya sesuai network masing-masing

Flow

1. Fetch a video feed from esp32
2. Detect cars using YOLOv7
3. YOLO script executute car detection
4. Error is reduced by merging duplicated boxes
5. Largest car is then parsed and expored with respective frame number
6. Execute number Plate detection using OCR
7. Numberplate storing and matching
8. If match, send a trigger to ESP32-CAM

Progress

• Live Video feed
• Yolo Car Detection
• Reduce duplicated boxes
• Car parsing
• Number plate OCR
• Number plate store and matching
• Sending Trigger

Source Used

• https://github.com/AlexeyAB/darknet