Ask HN: Image/video recognition, what's the status quo and application?

I'm just curious in this field, but don't know where to start to look at after a few tries on google. So perhaps I simply have not found the right terminologies to start.

The specific applications I'm interested in is motion recognition (e.g., finding out the form of athletes is more efficient).

My uneducated guess is there should be some way to recognize an image first and video recognition is to process a sequence of images.

Thanks in advance.