Parking Lot Management System

Download Report

Transcript Parking Lot Management System

PARKING LOT
MANAGEMENT
SYSTEM
Natasha Kholgade, Nientzu Kuan, Joseph Rodrigo
March 18, 2008
CE Senior Design Projects II
Rochester Institute of Technology
Agenda








System Overview
Operational Stages
User Interface
License Plate Recognition (LPR)
Override Controls
Hardware Components
Testing Methodology
Questions ?
System Overview
The objective of this project is to use license plate
recognition to manage the entry and exit of vehicles by
managing a database indexed by license plate numbers.
Laptop






License Plate
Get sensor signal
Operate camera
Image processing
GUIs
Database
Send gate open signal
SCI
This system is employed for
traffic that frequents New York
City. (NY, NJ, CT)
SERVO
SENSOR
Operational Stages
The system is started by switching on the PC and power supply to the microcontroller
and servo-based gates. Gates start off in the closed position.
Sensor placement
Activation of Sensor
A shows that a car is
present, and leads to
License Plate
Recognition and
permission to open
the gate
Activation and then
deactivation of
Sensor B signifies
that a car has
passed and left, so
the gate can be
closed
User Interface: Operator GUI
Image
captured by
camera
Database of
cars in the
parking lot*
Image of
license plate
Override
controls
* Format: ABC1234
NY
<start time>
<end time>
User Interface: Driver UI

Driver User Interface
 Two
boards with LEDs will be present: one in the entry
lane and one in the exit lane
 The lit-up LED will represent what action should the
driver take
LPR is successful
LPR is not successful
License Plate Recognition (LPR)
The following steps are involved in License
Plate Recognition




Object Detection
Plate Extraction
Character Segmentation
Character Recognition
License Plate Recognition (LPR)

License plate will be located with template matching by normalized
cross-correlation technique.


The three states will be searched and the state with the highest crosscorrelation index will be selected.
The character section will be extracted by moving a couple of pixels
below the state image.
0.6
0.4
0.2
License Plate Recognition (LPR)


Character segmentation will be done by horizontal
and vertical projections and image binarization.
Character recognition will be performed by passing
the segmented characters through a neural network.
Override Controls
These can be used by the operator in case of an exceptional circumstance, for
example, if the camera is down, or the sensors cannot detect a car passing before
them.
Override Command
Description
Suspend System
Suspends the operation of the sensor, camera, and LPR
process.
Restart System
Re-enables the operation of the sensor, camera, and LPR
process.
Open Gate
Turns the gate to the OPEN position.
Close Gate
Turns the gate to the CLOSE position.
NOTE: Override controls are specified individually for the entry and exit lanes.
Operation of SCI

The serial communication interface (SCI) will be used to send ASCII codes that will
control the functioning of different tasks
Function
Direction
Entry
Exit
lane code lane
code
Use sensor data to activate camera From microcontroller to PC
$41
$43
Use result of license plate
recognition to open gate
From PC to microcontroller
$42
$44
Override – suspend system
From PC to microcontroller
$45
$49
Override – restart system
From PC to microcontroller
$46
$4A
Override – manually open gate
From PC to microcontroller
$47
$4B
Override – manually close gate
From PC to microcontroller
$48
$4C
Hardware Components







2 Logitech QuickCam cameras
Laptop
4 OPB720B sensors
MC9S12DT256 microcontroller
2 Servo motor for gates
5 LEDs
SCI cable
Hardware Components: Camera




Logitech QuickCam camera
Still image capture: Up to
1.3 megapixel with
software enhancement
Up to1280x1024 pixels of
resolution
High quality VGA sensor
with RightLight™
Technology
Hardware Components: Sensor







Optek OPB720B
Long distance reflective type switch
Maximum detection range of 0.04” to 12”
Operating temperature of 0 to 70℃ (32 to 158℉)
Maximum collector current of 50mA
Low voltage output of 0.8V
High voltage output of 3.5V
Hardware Components: Microcontroller





16-bit HCS12 CPU
8-bit and 4-bit ports with interrupt functionality
Two asynchronous SCI ports
Enhanced Capture Timer
8 PWM Channels
Testing Methodology

Unit Tests

License plate recognition system:

Camera controls:

Sensor switch:

Serial communication interface:
Use variety of license plate images from NY, NJ and CT to ensure that the
developed software detects the text correctly
Call camera API in Visual C++ and make sure that it responds accordingly (i.e.
starts, stops and captures); adjust camera to have required resolution
Place car before and after borderline and observe if the appropriate responses
are obtained
Test send and receive functionalities from microcontroller’s viewpoint and
from PC’s viewpoint respectively
Testing Methodology

System Tests
A car comes into the entrance/exit lane, license plate
number is extracted, and car is allowed to pass
automatically
 A car comes into the entrance/exit lane but its license plate
number cannot be extracted
 Operator manually enters license plate number
 Operator manually opens gate
 Operator manually closes gate
 Operator manually suspends system
 Operator manually restarts system

Anticipated Problem Areas

Issues with License Plate Recognition software
 The
success of cross-correlation will depend upon how
close in size is the template of a state to the image in
the car
 Multi-scale
cross-correlation will have to be implemented for
cars of different sizes.

Sensor response
 The
response of the sensor depends upon the material:
different materials will generate different distance
thresholds. These will have to experimented with.
Questions