Find Areas That Are Too Thin In An Image But Aren't An End Of A Pointy Shape

This is a follow up question on my previous question. (Finding areas that are too thin using morphological opening on black and white images) After reading and implementing the sug

Solution 1:

You might want to consider using bwmorph operation 'endpoints' applied to 'skel' of your template - these two morphological operations should help you identify the the "pointy" shapes of your input image, thus excluding them from your "thin regions" you highlight.

Using opencv, you may find this example of morphological skeleton operation useful. It would also seems like pymorph can prove useful for you.