Optical Recognition Project [IMAGE PROCESSING METHOD]

Hi. i am dealing on OCR project and my task is to extract the darkest part of the image(which is my object) the common beginning process was, rgb->gray->image processing->binary

For image processing, i tried adjusting brightness method. when i display a histogram of it, value mainly were at the right side(white), my prof didn't see any at the left side(dark) which is the darkest side that we want. therefore, he says it doesn't seems a need to use adjusting brightness method. Can someone suggest a image processing method that i try?

thanks!

[Image Analyst]

Where did you upload your image?

[Adela]

this was the adjusted brightness's histogram.

[Image Analyst]

OK - those are the histograms, but where did you upload your original grayscale images?

[Adela]

I do not have the 'image' with me now, but i have the gray scale histogram.

[Image Analyst]

But aren't the numbers you are going to classify in your grayscale images? And I can't tell what's the best approach to background correction if all I have is the histogram.

[KIRAN kumar]

Try adaptive histogram equalization like adapthisteq().