Parking Lot App

36
Parking-Spot Locator Application Van Buren Boys Adrian A. Salazar Ajay Joshi Alex Humphries Charles Goodrum Michael Holloway Michael Koch Issued: April 27, 2012

description

Slide show of my Senior Design project.

Transcript of Parking Lot App

Page 1: Parking Lot App

Parking-SpotLocator Application

Van Buren Boys

Adrian A. Salazar

Ajay Joshi

Alex Humphries

Charles Goodrum

Michael Holloway

Michael Koch

Issued: April 27, 2012

Page 2: Parking Lot App

Welcome

Contents

• Section 1: Introduction to our product

• Section 2: The Mechanics of the Application

• Section 3: Development

• Section 4: Documentation and Maintenance

• Section 5: Marketing

• Section 6: Conclusion

2

Page 3: Parking Lot App

INTRODUCTION OF OUR PRODUCTSection 1

Page 4: Parking Lot App

Background

• We developed for the past year a system to aid commuters of the Discovery Park building to locate parking spaces for their vehicles.

• The idea is that the user would be able to access information about the status of the parking lot that they are visiting.

• Many parking lots across our nation suffer from problems of this nature.

4

Page 5: Parking Lot App

Goal

• We want people to know where they can park in the building and thus spend less time to search for it.

• Our product aims to implement a system using an image processing too to locate parking spaces and transmit this information in a simple format to the user via mobile application.

5

Page 6: Parking Lot App

THE MECHANICS OF THE APPLICATION

Section 2

Page 7: Parking Lot App

Components

• Camera takes images of the lot at appropriate resolution

• Image Processor detects parking spaces and places a .txt file of binary logic

• Server Retrieves, stores and sends data through querying tables

• Application reads tables and updates accordingly to the server communication.

• Digital Camera

• Go-Pro Hero 2

• Image Processor

• MATLAB

• Server

• Web host

• MySQL

• Mobile Application

• Android OS v. 2.3

7

Page 8: Parking Lot App

Web Host

Empty = 0Occupied = 1

T- 60000

<script/>

C Compiler

InternetConnection

8

Page 9: Parking Lot App

Camera

• The Go-Pro Hero 2Camera is characterized by its fast interval of picture taking.

• It is mounted on the roof of the Discovery Park building, aiming at the Parking Lot 2.

• It has a versatile rate of picture snapping that allows our system to have a fast rate of updating.

• Images saved on an SD card and moves to the image processing system for the next step.

9

Page 10: Parking Lot App

Image Processing

• The image arrives for processing in full color:

– The first step before doing anything is to grayscale the image

– Pin-cushion transform

• Flat the image, since it is being taken on a wide angle lens

– Edge Detection

– Histogram Detection

– Image Subtraction

10

Page 11: Parking Lot App

Edge DetectionLooks for edges in the parking spaces

Builds a ratio from the number or edge pixels

11

Page 12: Parking Lot App

Histogram Detection

The procedure for this test is to count the pixel intensities for number of occurrences.

12

Page 13: Parking Lot App

These are histogram plots from the previous slide’s images.

13

Page 14: Parking Lot App

Image subtraction

No change Change

14

Page 15: Parking Lot App

Server

• Acts as a data storage center

• Retrieves, stores and sends data accumulated from the image processor

• Programmed on MySQL

• Scripted tables with PHP to update

• Connects to both the image processor and the application

• On the web-server, it uses tables to refer to the parking spots

15

Page 16: Parking Lot App

Mobile Application

• Runs on Android OS v. 2.2

• Currently works for 4in screen phones

• The application uses a very simplistic screen to display the results of our camera’s image processor.

16

Page 17: Parking Lot App

APPLICATION DEMO VIDEO

This is a sample demonstration of our product in action. Note that for purposes of testing and resources, we only implemented the system for one parking lot.

17

Page 18: Parking Lot App

THE DEVELOPMENTSection 3

Page 19: Parking Lot App

Timeline (August 2011- December 2011)

2011 Van Buren Boys Project

Team Formed

DecemberOctoberSeptemberAug November

Project Plan

Requirements Document

Preliminary Design Document

Detailed Design Document

Revise Project Plan...

Project Presentation

Project Presentation

Materials List

19

Page 20: Parking Lot App

Timeline (January 2012- April 2012)

2011 Van Buren Boys Project

Revised Plan

MayMarchFebruaryJanuary April

Implementation Plan Document Project Presentation

Test Plan Document

Final User’s Guide Document

Final Maintenance Manual Document

GoPro Camera Delivered

Final Acceptance Test Document

Android SDK Downloaded

Server Host Running

Camera Installation and Image Proccessor Coding

Primal Function of Application Application-Server Connection

Appliation’s HCI change

Launch App. To the Market

Image Proccessor ALgorithm

Final Report

Succesful Connection

20

Page 21: Parking Lot App

Testing & Results

Camera

• Snaps pictures at a good rate

• Weather proofed

• Easy to mount and dismount

• Sends pictures immediately

Image Processor

• Car detection with more than 90% accuracy

• Vacancy detection

• Dataset creation for the server to use

21

Page 22: Parking Lot App

Testing & Results

Server

• Communicates with the application

• Updates done automatically through scripting

• Tables are easy to read through

Application

• Runs properly on 4in screen phones with Android 2.2

• Catches errors correctly

• Opens and closes correctly

• Responds to User Interaction

222

Page 23: Parking Lot App

DOCUMENTATION AND MAINTENANCE

Section 4

Page 24: Parking Lot App

System Interaction

Image processor

24

Page 25: Parking Lot App

Class Diagram

25

Page 26: Parking Lot App

Package Diagram

GoPro_Package

SD_CARD

26

Page 27: Parking Lot App

Components Diagram

GoPro_Package

27

Page 28: Parking Lot App

Maintenance

Camera and Image processor

• The camera needs to be cleaned and adjusted every so often so images remain accurate

• Image processor must have a constant cleanup of files as many files will expire or are too old.

Server and Application

• Server must be up and run 24/7 and that information is always accurate.

• Web services must maintain a its fee payment.

• Application must maintain with time and versions of phones and code.

28

Page 29: Parking Lot App

MARKETINGSection 5

Page 30: Parking Lot App

Our Client and Market

• We aim to release our product to the students, staff and faculty members of Discovery Park to test it

• We also consider to expand our idea throughout all the building and maybe go even further by implementing the idea for the entire campus of UNT.

• We can go even further by selling the ideas to developing groups or companies, but that would be a decision for another day.

30

Page 31: Parking Lot App

Android Market

• We plan to release this product through the Android Market and Google Play in May, so people can start using the application and test it over the summer to check for any malfunctions or bugs.

• Its use will be official for the next semester of Fall 2012.

31

Page 32: Parking Lot App

Future Considerations

• Certain considerations to the system have been thought

• We aim to increase the capacity of our technology all across the UNT Campus.

• We also consider to develop the application for iPhones or any iOSsystem.

• Finally, we think about making the product a profitable application

32

Page 33: Parking Lot App

CONCLUSIONSSection 6

Page 34: Parking Lot App

Review of Product

• Our product is an Android Application for standard mobile phones of 4in screens and run the OS v. 2.2

• It aims to aid commuters of the Discovery Park building to locate parking spaces for their vehicles.

• The idea is that the user would be able to access information about the status of the parking lot that they are visiting at any time.

34

Page 35: Parking Lot App

Review of System

• The system follows 5 easy steps

– A camera snaps a picture

– Picture is sent to the image processor and makes binary values to determine occupancy

– Values are stored to a server

– The application will try to talk to the server as often as possible and update a simple representation of the parking lot.

– Rinse and repeat

35

Page 36: Parking Lot App

Q&A SESSION