object recognition in an image using python

We use Python (WinPython 3) and OpenCV for image processing programming.

What should I do to make the building in the image recognized in the image taken with a digital camera?

Specifically, if there are several buildings lined up against the blue sky, I would like to recognize all the structures that overlap the sky as objects.Also, I would like to be able to extract the intersection point between the graph and the building when overlapping the graph on the image.

If you know more about image processing, please let me know the procedure and method in an easy-to-understand way.Thank you.
(The red circle in the image recognized)