I believe it's a combined process, i.e image recognition is for finding objects and meta data extraction is to extract GPS location. I don't think the links in

I believe it's a combined process, i.e image recognition is for finding objects and meta data extraction is to extract GPS location. I don't think the links in the JSON image will help "the criminals" :)