A Python based Computer Vision Project for Car Parking Operations System
PyParkOps automates parking lot management through machine learning and Bangla license plate recognition. Vehicles are detected in real time, plates are localized and digitized, and parking fees are calculated automatically while operators monitor everything through a Streamlit dashboard.
- deliver an end-to-end Python ML workflow
- employ YOLO object detection for live inference
- run Bangla OCR for plate transcription
- ship a Streamlit dashboard tailored to parking attendants
- eliminate manual parking fee computation
- YOLOv8 for multi-class vehicle detection
- Custom/finetuned YOLO for Bangla license plate localization
- EasyOCR to interpret Bangla characters
- OpenCV for frame capture and preprocessing
- Pandas as the logbook and fee calculator
- Streamlit for live dashboards and controls
- Webcam streams frames into the pipeline.
- YOLOv8 marks vehicles; plate model crops license regions.
- EasyOCR extracts Bangla plate numbers.
- Entry time and plate data land in a Pandas dataframe.
- On exit, duration and fees are computed automatically.
- Streamlit displays the live feed, history, and earnings metrics.
Refer to the flow diagram in the project report for a visual breakdown (Project Report OOP).
- real-time vehicle detection overlay
- Bangla plate OCR with accuracy validation
- automatic entry/exit timestamping
- rule-based parking fee calculator
- live Streamlit dashboard with camera preview
- daily earnings summaries and CSV export
PyParkOps minimizes manual errors in Bangladeshi parking lots by merging computer vision, OCR, and web dashboards within one deployable solution.