View
1
Download
0
Category
Preview:
Citation preview
Slide 1
SolutionsHYSTER COUNTING
APRIL 2007Presented byBarry T. Fryer DudleyBarry T. Fryer DudleyBarry T. Fryer DudleyBarry T. Fryer Dudley
(MBA {IT}; MSc {Image Analysis}
“..any sufficiently advanced technology is indistinguishable from magic.” Arthur C. Clark
Quinn GreenOperations Manager KZNquinn.green@itoitechnologies.com0824950935031 -2420500
Slide 2
SolutionsRequirement: Requirement:
Counting hysers with high Counting hysers with high accuracy (over 99.98%).accuracy (over 99.98%).
Trend analysis Trend analysis -- minute by minute minute by minute update, recording all hysers into update, recording all hysers into and out of the pens.and out of the pens.
Slide 3
SolutionsWhat is Hyster Recognition?Hyster Recognition is an image
processing technology used to identify hysters by a unique number. It is a non-intrusive, computerised method of matching a hyser numberto a database of numbers, allowing accuracte counting to occur within 300ms (@ 99.98%).
Slide 4
SolutionsA Word About Our Eyes
Eyes are very good contrast adjusters, but not good for distinguishing subtle variations in colorEyes can discern about 30 continuous levels of gray or color in a field of viewEyes are not good judges of distanceEyes cannot accurately reproduce measurementsEyes can not work in the dark or 24/7
Slide 5
SolutionsSAME SIZE???
Slide 6
Solutions
Why do Image Analysis?Improved Precision /Accuracy in MeasurementsReproducibility of ResultsHigher Throughput than Manual MethodsBetter Definition of Contrasting AreasMore Measurements / FasterReal Time Link to Databases
Vehicle:
• Size
• Colour
• Shape
• Texture
• Grey level
Slide 7
Solutions
1 - Capture2 – Find unique number
3 – OCR
4 – Report & Count
Hyster Counting - 5 easy steps:
5 – Alarm
Slide 8
Solutions
Camera/IlluminationUnits
(Up to 6 per system)
Frame Grabber
Hyster Software + DLL+Sample Client Program
PC Station
Included HardwareI/O Card +Terminal Block
Power Supply for SCH
ENTRY (X no)
EXIT (X no)
Slide 9
SolutionsSuggested camera is a black & white optimized camera with high resolution. Illumination unit is integrated near by the camera and designed to assemble as one assembly. This unit camera + Illumination is supported by the SEELANE application and utilizes its additional features to maximize the results. The application controls the level of illumination (4 levels) and the image capture is synchronized with the illumination to achieve higher signal to noise ratios.
SEELANE
Slide 10
Solutions
Lens
Image sensor
Digitizing Pre-processing
Timing Interface
PLC
Cat-5eEthernet cableup to 100 m
Local I/Os:-Trigger input-Results output
Illumination control
Image Processing
in
PC
Illumination
(Lens IrisVideo)
Power
Camera and Illumination system
Slide 11
SolutionsImage –• A stand alone, human readable monochrome JPEG image with
approximately 768 pixels by 288 (CCTV cameras). • This image will display the detected plate of the best
recognized within the set of images that are captured for that event.
• For driver or load ID an additional color jpg image 768 by 288 pixels, used for optional review. It will share the same file name as the plate image, but will have prefix letter.
Slide 12
SolutionsSuggested label to uniquely ID hysterSuggested label to uniquely ID hyster
Suggested, reflective plates the system is designed to operate withSuggested, reflective plates the system is designed to operate with
Slide 13
SolutionsSuggested position Suggested position ffor hyster IDor hyster ID
Suggested position for the hyser IDSuggested position for the hyser ID
Slide 14
SolutionsSuggested position Suggested position ffor hyster IDor hyster ID
Slide 15
SolutionsSuggested position Suggested position ffor hyster IDor hyster ID
Slide 16
SolutionsSuggested position Suggested position ffor hyster IDor hyster ID
Slide 17
Solutions
Camera/Illumination unit for access-control applications
B&W camera with mounted lens
Pulsed LED array (Near Infra-Red or Red spectrums)
Weatherproof Case: IP 65 , Enforced Poly-Carbonate, UV protected
Control circuit (sync and pulse control, 4-levels illumination control)
2-Degrees freedom Mechanical interface
Low energy consumption (15VDC 3A)
Slide 18
Solutions
Slide 19
Solutions
Pulsed Illumination for 24 HR operation
Slide 20
SolutionsSchematic Diagram
SEELANE
Slide 21
Distance and camera angles for moving vehicle:Front and Rear Fixed Cameras
Camera Illumination units
5 000mm 5 000mm
Slide 22
SolutionsPossible distance and camera angles mounted
above the hyster over the entrance / exit
Slide 23
SolutionsHyster Recognition for entry (1 of 4)
Hyster approaches entry point (with or without boom)
Slide 24
Solutions
Camera /Illumination unit
LPR
MKF 359 GP
Vehicle Sensor
Recognize plate number a compare to
data base
PC Station
Hyster Installation for entry (2 of 4)
Slide 25
Solutions
Recognition Rate: 0.1 sec
Options exist to capture driverinformation OR link driver/vehicle load to plate
Picture of load could be captured, possibly
automatically identified
LOAD
captured
(3 of 4)
Slide 26
Solutions
Total time required for an authorized vehicle (correctload, face and hysterplate): 1-4 seconds
If hyster allowed in via database, boom opens, IF NOT, guard calledto investigate
(4 of 4)
Slide 27
SolutionsPossible Matching Criteria
3 drivers of 20
0987654321
123456789
NP132785
Result
OVERALL – After hours, hence not allowed!
AllowedHyster detailsAllowed to drive hyster
AllowedFacial Capture
Allowed123456789Staff No
AllowedHyster Recognition
ResultCapturedCriteria
Slide 28
SolutionsEXIT PROCEDURE:
•Hyster recognition occurs as vehicle approachs exit via pulsed IR camera,
•Image capture of driver / load if required via colour camera,
•Recognition occurs in 100-300ms
•Database updated and count increased;
Slide 29
SolutionsSoftware: How to enrol hysters:
Slide 30
SolutionsLPR LOGLOG FOR A SINGLE EVENTAdded,1,NP192785,100,100,100,100,100,100,100,100,100
Added,2,NP132785,100,100,100,100,100,100,100,100,100
Added,3,NP12785,91,100,100,75,100,100,100,100
Added,4,NP12785,91,100,100,75,100,100,100,100
Best,2
Result,NP132785
Added,1,NP132785,95,100,100,100,100,100,100,100,100
Added,2,NP132785,100,100,100,100,100,100,100,100,100
Added,3,NP132785,100,100,100,100,100,100,100,100,100
Added,4,NP12785,96,100,100,75,100,100,100,100
Best,3
Deleted,4
Result,NP132785
Added,1,NU44667,99,100,100,100,100,100,100,100
Added,2,NU44667,99,100,100,100,100,100,100,100
Added,3,NU44667,99,100,100,100,100,100,100,100
Added,4,NU44667,99,100,100,100,100,100,100,100
Added,5,NU44667,99,100,100,100,100,100,100,100
Added,6,NU446M7,75,100,100,100,100,30,40,100
Best,3
Deleted,6
Result,NU44667
HYSTER PLATE, LANE, USER, DATEHFP848GP,..........,1,Yes,Rachel Tim,Sat Aug 14 09:05:16 2004,
KBW141GP,..........,2,No,,Sat Aug 14 09:05:18 2004,
MKF359GP,..........,1,Yes,David Smith,Sat Aug 14 09:05:20 2004,
LJD977GP,..........,3,No,,Sat Aug 14 09:05:22 2004,
ND474276,..........,2,Yes,Daniel Bush,Sat Aug 14 09:05:24 2004,
DYK370GP,..........,3,No,,Sat Aug 14 09:05:30 2004,
LGW487GP,..........,1,Yes,Goldy Luck,Sat Aug 14 09:05:36 2004,
KPR835GP,..........,3,No,,Sat Aug 14 09:05:38 2004,
All images are kept
Slide 31
SolutionsEntry
Exit
Slide 32
Solutions
• The systems employs the SeeCar OCR engine, which will run on the local processing units. The OCR engine processes images, locates the relevant plate ID in the image, and produces an alphanumeric result for each image processed. The OCR engine is based on neural network technology and can be trained to recognize different fonts, characters and syntax. The systems supplied for this Project are specially trained to recognize license plates suggested.
SeeCar OCR engine
Slide 33
Solutions
Database remotely updated
Automatic Database updated
Trigger when a plate seen
Configuration
Slide 34
SolutionsThe data will be transmitted to the TCS in two forms:
• Windows DDE (Dynamic Data Exchange) Message - sent to the CCS server TCPIP network. The DDE will contain the VES Optical Character Recognition described below.
• Image file - which will be stored on the VES server, then transmitted to the NCSvia a dedicated transfer service running on the Trip Processing Server.
Slide 35
Solutions2) VES Optical Character Recognition Data:
• PC unique ID integer number• License Plate string• TCS Synchronization Code • Date and Time of Image Capture• File Name ( a link to the name of the resulting .jpg file
stored in the VES, plus the overview image if present)• Confidence of the recognition result• Plate flag• GPS data (if present)• Other data as required
Slide 36
SolutionsData Analysis Data Analysis –– LPR successLPR success
Slide 37
SolutionsRemote Control Remote Control Controlling a ComputerControlling a Computer
The remote computer’s desktop is
displayed in a window on the local computer.
Multiple computers can be accessed
simultaneously.
The remote computer’s desktop is
displayed in a window on the local computer.
Multiple computers can be accessed
simultaneously.
>>
Slide 38
Solutions
BAKER4ZN,..........,1,No,,Thu Jul 27 18:01:45 2006,C:\Program Files\HI-TECH Solutions\SeeWay\New E\060727\11_15479750_1.jpg
DATA is sent in REAL TIME via TCP/IP
Slide 39
SolutionsReal Time Plate Identification
TRZ235GP,..........,1,No,,Thu Jul 27 18:01:42 2006,C:\Program Files\HI-TECH Solutions\SeeWay\New E\060727\11_15477578_1.jpg
BAKER4ZN,..........,1,No,,Thu Jul 27 18:01:45 2006,C:\Program Files\HI-TECH Solutions\SeeWay\New E\060727\11_15479750_1.jpg
PFJ654GP,..........,1,No,,Thu Jul 27 18:01:46 2006,C:\Program Files\HI-TECH Solutions\SeeWay\New E\060727\11_15481343_1.jpg
NP74555,..........,1,No,,Thu Jul 27 18:01:47 2006,C:\Program Files\HI-TECH Solutions\SeeWay\New E\060727\11_15482343_1.jpg
Slide 40
SolutionsDaily Report: To be customised
Slide 41
SolutionsShift Report: To be customised
Slide 42
Solutions
Slide 43
Solutions
Slide 44
Solutions
REQUEST A LIVE DEMO
Recommended