Super Smart Parking

Prerequisite

• Python 2
• OpenCV 3.4.1
• Sklearn, shapely (Python 2)
• Java
• Gearman Service: Gearman Job Server, Gearman Worker (python 2), Gearman Client (Java).

Run Project

• Run detect plate worker: Go to detectPlate folder and run: python detectPlate.py
• Run SmartParkingWeb:

• Config Apache Tomcat with NetBean.
• Open Netbean Project and run.

Project Detail

1. Detect Plate System:

• OpenCV, SVM, Gearman Worker.
• Data:

• GreenParking 1750 item
• BSXM-Google 657 item => Filter => 184 item ==> Sum = 1934 item

• Generate Character: 11.702 item => Filter => 11275 item

• 9618 item number / 10 labels
• 1608 item char / 21 labels => Augmentor data => 17.037 (Char augmentor)

• Detect Plate by OpenCV
• Recognize character in plate by Machine Learning
• Machine Learning:

• SVM Multiclass Kenel RBF.
• Number: C=32, g=0.000112 => Accurancy: 98.32%
• Char : C=32, g=0.0078125 => Accurancy: 99.87%

2. Web site system:

• Backend: Java web
• Front-end: HTML, CSS, JS
• Spring Framework with MVC
• Gearman job with ML SYSTEM.

Contact me

• Facebook: fb.com/linhpph
• Email: linhphanhust@gmail.com