Sort:  

Indeed... even Google with their seemingly endless resources hasn't gotten down to it in a meaningful way.

I've played with image recognition a bit but just wondering, what makes it so costly at the moment? The storing of the images or running the image recognition code?

One would need to purchase API calls for reverse image search. I haven't been able to find a reasonable one. :/

I see... thanks for explaining. And i guess there's no way to know how good those APIs/ index is in the first place.