2
$\begingroup$

I am studying computer vision for the past 3 months. I have come across the object identification problem, where given an image, CV would identify various parts in the image.

If I give an image, and a rectangle coordinates, can CV identify the parts' names within that rectangle? For example, can I train a model to identify the parts in the below image (mountain, river, in this case)? The model should not identify other parts like flowers, sky, etc., as they come outside the rectangle).

I tried searching but could not find similar problems. Can anyone give me a direction to solve this problem?

enter image description here

$\endgroup$
0
0
$\begingroup$

Giving this small rectangle to a model will probably not provide him enough information to identify what is inside this rectangle. An approach that could work would be to do image segmentation, having the identity of everything in the image, and then extract the label of the pixels in the rectangle.

$\endgroup$
1
  • $\begingroup$ Yeah. Or crop the region and compute only for the region of interest. This can speed up. $\endgroup$ Jun 19 at 11:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.