OCR Character set

[co2junkie_]

greetings,

I am very new to machine vision, I am trying to create a number plate recognition system, i have trained each set of character in number plate in OCR Training, and when im opening each individual character image, its successfully converts it into string, but to a whole image (of number plate) its not responding ( i am getting '?' ) i have tried using ROI for specific area selection but doesnt work.. I don;t know what seem to be the problem please help.

[ChristopheC · Accepted solution]

You need to connect the Region of Interest to the IMAQ OCR Read Text 3 vi.

The following example shows on how to use the OCR VIs:

C:\Program Files\National Instruments\LabVIEW\examples\Vision\2. Functions\OCR\OCR First Example.llb\OCR First Example.vi.

With this example, you can even load your character set file and try it out on your images.

Hope this helps.

Best regards,

Christophe

[co2junkie_]

Thank you very much sir, its working fine now.. 😄

[Er.DKD]

hello

I am new to labview and i dont know how to train characters.actually I am doing project on ocr system and i have tried somewhat but i am unable to understand what to do next.please help me.please provide me any video tutorial which can help me easily.

[The_Hunt]

Hello Dilpreet,

The Vision Development Module ships with a couple of Optical Character Recognition (OCR) examples that provide as a great platform to begin your application. You can find these by going to the LabVIEW Help >> Find Examples >> Toolkits and Modules >> Vision

Inside there are 2 OCR examples: 

Functions >> OCR >> OCR First Example.llb

Applications >> OCR Example.llb

You can also use the OCR Training Interface program, which is located under Start >> Programs >> National Instruments >> Vision >> OCR Training. Then click on the Help and select Tutorial. 

I would also highly encourage you to read through the NI Vision Concepts Manual, particularly Chapter 18. 

There are several ways we can work with training OCR within LabVIEW. The easiest way is to use the NI Vision Assistant (Function palette >> Vision and Motion >> Vision Express >> Vision Assistant). Here you will find several processing functions. Under Processing Functions: Identification you will find OCR/OCV. If you select that option, you can either set (under the 'Train' tab) a Character Set Path or create a New Character Set File... If you select the latter it will open the NI OCR Training Interface. This tool is used to create your own character set file which you will be able to use for your OCR operations. There are instructions included with the interface which will assist you with the process.

[Er.DKD]

Thank you so much sir

I'll definately try all this

[Er.DKD]

Hi co2junkie_

can u do me a favour please provide me help regarding OCR and please if u can send me your vi for OCR

Thanks in advance