Automatic Vehicle Entry Exit Log - Senior Year Project

Post on 13-Apr-2017

125 views 6 download

Transcript of Automatic Vehicle Entry Exit Log - Senior Year Project

AUTOMATIC VEHICLE

ENTRY/EXIT LOG Ahmad AzharUmer Numan Guraya Umar Tabish Haque Taiyaba Saeed

AdvisorDr. Jahangir Ikram

Co-Advisor:Dr. Tariq JadoonDr. Murtaza Taj

Project statement:The motivation for our project stems from a demand in the market to develop a system which can:

• Maintain an online log for entry/exit time of vehicles in a certain premises (factory etc)

• Maintain an online record with pictures for vehicles entering/exiting the premises

• Extract the number from these pictures through ANPR(Automatic Number Plate Recognition)

• ANPR should work on all legal number plates within Pakistan. E.g. Lahore, Karachi and Islamabad number plates

• Provide complete record of vehicles inside a premises at any given time, time of entry and exit of vehicle, setting up an action upon entry or exit of a vehicle (SMS etc – future additions)

• Provide vehicle count for past 10, 30 and 60 minutes.

Aim of Project• Maintains a data for the entry/exit time of vehicles.

• Solves administrative issues by generating a text log for vehicles entering/ exiting the premises. This helps keep track of vehicle inflow, outflow and no. of vehicles in the premises which is helpful in fleet management.

• The system has a provision of keeping a record for a given period of time (e.g. 1 month).

• The system allows the user to ascertain the status of vehicles online, to know whether the vehicle is present in the premises or not.

• Provides a fast, efficient and reliable method of identifying presence of vehicles in a premises compared to traditional methods

HOW THE SYSTEM WORKS

Sends Signal

Trigger Camera

Send Image to Raspberry Pi

To ANPR System

Extracted Number

Info to WebsiteTo

server

“AVELOG” is divided into three main parts

• Vehicle detection

• Database development and online log

• Number plate Identification via ANPR

Vehicle Detection

Vehicle Detection

Loop Specifications

1. 40/76 wire2. 4 turns3. Rectangular Octagon4. Twisted cable

Vehicle Detection – the circuit

Vehicle Detection – loop response• Change in voltage transmitted as car passes over

Vehicle Detection• Voltage change ~0.2V

• Amplification of voltage change using ADC

• Setting upper and lower threshold

• Extensive testing to ensure no false readings

𝑈𝑝𝑝𝑒𝑟 h h𝑇 𝑟𝑒𝑠 𝑜𝑙𝑑❑

Image Capture

• Image initially saved in RaspberryPI

• Image later transferred to Server using SSH communication

• Automatically run code every time RaspberryPI restarts using crontab

• Image name taken to be date and time i.e. 2016-04-22.09:00:00.jpeg

SSH Communication

• Assign a static IP to both devices

• Use keyless SSH entry for communication between RaspberryPI and Server

• Generate public keys for both devices and share public keys among them

• Using crontab, send all images saved in RaspberryPI to Server every minute

• Delete image from RaspberryPI after sending image to Server

• Allows user access to Raspberry PI from Server

Website Development• Images recived in html folder of Server, where website is being hosted

• Installation of following softwares:• Apache• PHP• html• Mysql• Phpmyadmin; GUI for databases• Nodejs

• traffic.lums.edu.pk

Website Development – Main Page

Website Development – 10 min

Website Development – 60 min

Website Development – No image

Website Development

• Storing name directly into database

• Breaking name to get time and date and storing it into database

• Make four variables:

• Start_time; gives initial time

• end_time; gives end time

• date_start;

• date_end;

Database

ANPR using WebcamImage from Current Position Image from Up Close

ANPR using WebcamGray Scale Image

ANPR using WebcamBinary image

ANPR using WebcamBounding Boxes plotted around the connected components

ANPR using Webcam

LED 7658 LE 5033

Objects Extraction using Template Matching

ANPR using Phone CameraImage from Current Position Image from Up Close

ANPR using Phone CameraIdentification of the Connected Components

ANPR using Phone Camera

LEB 8936 LEB 9493

Objects Extraction using Template Matching

ANPR- ConclusionThe analysis on the mentioned methods leads to following conclusion:

• The camera must be of high quality

1.3MP Night Vision Waterproof License ANPR License Plate Capture(LPR) 1080P IP Camera

Plate Recognition Camera

US $ 275.00 US $255.00

ANPR- Conclusion• Camera Positioning

• Ideal position is at road level

• Single camera covering one lane

• Distance of the camera from the vehicle should be kept minimum

Hardware InstallationThe system was deployed at the out gate of LUMS. The following were included in it:

1. Webcam 2. Raspberry Pi (controller/client)3. Loop controlling circuit

Cost Analysis

ITEM COST(PKR)

Raspberry Pi 5000

Webcam 8000

Wire (for loop) 1000

Labor 700

Hardware 1600

Total 16300

ETAG and our system • No initial cost and logistics required to paste sticker on every vehicle

• No need to bring all vehicles within the system for system to work

Future Work

• The proposed design is a prototype of cost effective solution to address an

important administrative issue. Potential future work on this project

includes:

• Application of same system on the entrance of LUMS.

• Improved image processing techniques for number plate identification

• An SMS alert system to inform arrival/exit of vehicle of choice