What is the best way to use projection histograms as characteristics in classifiers or Nns without using a resize on the image

Asked

Viewed 107 times

0

I’m doing some work for digit recognition on digital displays and I’m looking to use projection histograms as characteristics for some classifier, however the images are all of different sizes resulting in a different number of features for each image but I need an equal number. I am working with Opencv in C++, I thought to use the resize function, but this would leave the distorted images, I would like to know a viable option to use the projection histograms as features circumventing this problem.

  • 1

    I believe your question is overly broad and based on opinions, doubly outside the scope of the site. Can you [Edit] the question and elaborate it better to avoid closing? Can you read [Ask] for more details.

  • I don’t understand, I have a problem and I’m asking some way to solve it, since I have no idea. This is not completely within the scope of the site???!!!

  • 1

    The only thing you provided in the question was the statement of a problem. Making it seem like [en.so] is a technical support site to solve problems you couldn’t solve yourself. As clear as what your goal is, hardly a question will answer exactly what you are looking for, so it will be better if you [Dit] the question and present more details of the problem. You can give examples of what the different program entries are, how you plan to use the histograms, define which language you will use, etc.

  • Thank you, friend. Edited question, any more suggestions?

No answers

Browser other questions tagged

You are not signed in. Login or sign up in order to post.