I've built a handful of CNN using tensorflow, keras, pytorch for recognizing text/number/objects in an image. What I'm trying to figure out how to do now is how to recognize numbers on a piece of paper that are circled by a pen or pencil. So on a piece of paper, there are two rows of 6 sets of numbers, and someone will go through and 2-3 sets of numbers, and I'd like to be able to take a picture or scan the paper, and have a model be able to recognize which number sets were circled. What would be the best way to start going about this?
$\begingroup$
$\endgroup$
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
The problem you are describing is known as the Object Localization problem. The main DNN architecture used today includes R-CNN and related algorithms as well as YOLO algorithm. It's a well-studied problem in computer vision. you can found plenty of information online.
