Download presentation
Presentation is loading. Please wait.
Published byAnthony Owen Modified over 8 years ago
1
Automatic License Plate Recognition for Electronic Payment system Chiu Wing Cheung14036487d
2
Outline Introduction Theories Methodology Results Applications Conclusion
3
Overview License plate recognition is the extraction of license plate information from camera It makes use of the technology of image processing, optical character recognition It plays an important role in Intelligent Transport System
4
Objectives To Study different algorithms of image processing To develop an android ANPR application using OpenCV and Tesseract To build a model to demonstrate the operations after the license plates are recognised
5
Theories Automatic Number Plate Recognition (ANPR) comprises of six stages
6
Image acquisition Camera is essential element of ANPR system Camera is used to capture image of vehicle containing license plate information Camera can be triggered by: 1.Hardware 2.Software 3.Free Flow
7
Image acquisition Hardware: Physical sensor like magnetic loop detector It is used to detect the existence of vehicles Software Movement detection Using several algorithms like background subtraction Free Flow Detect license plate continuously
8
Image Pre-processing Image capture need pre-processing To reduce the noise of background To sharpen plate information To enhance the processing speed of recognition
9
Plate localization Find the region of license plate as Region of Interest Eliminate the unwanted background Many algorithms for different features of license: Edge information Texture feature Colour feature
10
Character Segmentation Identify the contours of each character Distinguish each character by Image Scissoring algorithm To normalize characters Enhance the accuracy of Optical Character Recognition
11
Optical Character Recognition Generate the plate information from input image The basic process: 1.Samples of characters are input for training in advance 2.Input image of segment characters 3.Comparing input image with trained data 4.output the results
12
Methodology Android application development OpenCV Tesseract OCR Arduino
13
Android App Development Android is a Linux-kernelled mobile operating system Most Android phones contain cameras, CPU, Bluetooth and Wifi Android software development kit (SDK) provides platform for developers to build their own application Eclipse used as a IDE for develop ANPR application
14
OpenCV Open source library for computer vision Multiple interface like C++, C, and Python Provides thousands algorithms for image processing
15
Tesseract OCR Open source OCR engine sponsored by Google Provide trained data of different languages
16
System design
17
Results
18
Vertical Sobel Operator
19
Close Operation
20
Plate Localization
21
Real test
22
Real Test
23
OCR
24
Application Electronic Toll Payment Traffic Surveillance Parking Management System Traffic Law enforcement
25
Conclusion ANPR application has been developed on android phone making use of open source libraries. Different algorithms for image acquisition, image processing, plate detection, and character recognition has been implemented
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.