Can someone help me in implementing image processing library of TI on ekk-lm4f232 evaluation board?? just want to know whether we can do or not?? and any code example if i can get

PRATEEK ARORA1 said:

i am trying to implement image processing on a ekk-lm4f232 evaluation board

Do you have any experience in "image processing"?

Have you done any background study into the subject?

PRATEEK ARORA1 said:

can someone tell me whether it can be done

That rather depends on what, exactly, you mean by, "image processing" - doesn't it?!

• What type of "image"?
• What type of "processing"?
• What speed do you require?
• etc, etc...?

PRATEEK ARORA1 said:

if i can get any code example for that??

http://www.lmgtfy.com?q=image+processing+library

thnx for reply

i basically want to do character recognization on ekk-lm4f232 .....

yes, i have some experience of image processing....but that was using matlab....

i want to write a c code using TI imglib and implement it on above mentioned kit...can it be done??

PRATEEK ARORA1 said:

implement image processing

Mon ami - is not your "specification" for such processing - perhaps a tad imprecise/lacking? 

But of course you can - should you not mind image flicker, minimal number of colors, and a, "Texas-sized" add-on pcb(s) to hold a multitude of much needed memory and other support components. 

Image processing clearly was not the intent when Cortex M-3/M-4 were conceived/implemented.  Indeed M-3 & M-4 serve a great need - have their place - but that is not, "image processing."

You may wish to dial-up TI's OMAP - especially the new OMAP 4470.   This is full-fledged, 4th generation, SoC - especially targeting such "image" applications.  So much of what you "appear" to seek plays on, "that channel" - not here...

thnx for the replies..

i am a newbie....

i don't have any other option than using LX4F232H5QD which is on EKK-LM4F232 eval kit ...

i want to do character recognizing on this...

i was thinking of writing a C program using TI IMGLIB and run it on tha eval kit....is it possible???

thnx

PRATEEK ARORA1 said:

don't have any other option than using LX4F232H5QD which is on EKK-LM4F232 eval kit

Presumably because this is a school project?

PRATEEK ARORA1 said:

i want to do character recognizing

Well Optical Character Recognition ("OCR") has been around for many decades, so there should be plenty of material available to study - this is your first step.

That study will give you an understanding of what is involved in doing OCR.

With that understanding, you will be able to evaluate whether the capabilities of the EKK-LM4F232 and TI IMGLIB are adequate...

PRATEEK ARORA1 said:

is it possible???

That's just one of many questions one may ask...  Perhaps more powerful is why you, "have no other options?"

Fitness & suitability of device to intended purpose usually rise to higher consideration in most "real" projects...

that's because i am taking part in TIADC and i have ordered this evaluation kit(got a suggestion from a friend) but he is facing a problem....

i cannot order anything else

i just want minimal info just to start....

Andy Neil said:

i am taking part in TIADC

Do you mean this http://www.uniti.in/adc ? [/quote]

So this is your project, then: 

7-1

A

Conversion of text document into Braille

BITS-PILANI,GOA Campus

Saurabh Bhatia, Prateek Arora

Sheron Figardo

http://www.uniti.in/adc/adc/101

How were you planning to share the prize money with helpers on this forum...?

 

 

hey.....

i did not ask for any code neither i asked for the solution of the problem...

i just asked whether i can solve the problem the way i was telling in the above posts....

that's it....

so don't post such comments due to which TI authorities might think that i am cheating or something...

thnx anyways

thnx very much...this was exactly what i needed.....

sorry it is not about sharing prize money but it is about doing things honestly so that we can learn...

even iu give me ur code for free i won' t take it.......

well anyways thn for the help

It warms my heart to read those words....

So for example:

http://www.abbyy.com/mobileocr/android/

and:

http://www.designnews.com/lecture-calendar.asp

I made a post with several links which should help. Hopefully a moderator will approve the post.

Best wishes.

In case you do not see my other post come through -- google these phrases:

ABBYY Mobile OCR Engine Arm Processor

smalllquiang/OCR github

After you OCR, you will need to parse and then analyze for "correctness" (Since OCR is never perfect -- you have mistakes.)  -- this is where you need to analyze symbol context -- hope that is clear. Variable context parser/compilers are considered a challenge at the grad student level -- they are possible despite the warnings in texts that grammar must be a "fixed grammar".

If you need this explained, I will try. If you cannot solve this problem, there will be "unfixable" errors in your output from the OCR phase.

Yes I really have done this... ;-)

Good luck.