excuse me. I am learning OpenCV and Visual Studio (c++).
I want to create an xml file that recognizes the landmarks captured by the camera (movie).
How can I create a high-precision xml file that can do that?

why you think, it would be an xml file ?

explain “landmarks”, please ? from geography ? faces ? poses/skeleton ?

if you HAD done ANY rsearch on this before, you probably would cite some related approaches, so ?

Uploading: WhiteMask.jpg…
I’m going to use the white side of the image as a landmark.

I want to create a file like the sample “haarcascade_frontalcatface.xml” to recognize the image.

please try again with the image …

This one.

  • if you haven’t already, please read the tutorial
  • you’ll need the opencv3.4 branch (the tools are no more in 4.x)
  • due to the highly reflective material, you will need a ton of positive example images (with varying lighting)
  • please note, that while haarcascades were a good thing, 20years ago, anyone else is using deep learning / cnn’s for this task now. again, please up on reasearch !

I will try.
Thank you very much.