Open afoeder opened 4 years ago
Hey @afoeder. Thanks for the interest. Could you please explain why would you like to do this? Yes, it can be done, however will lead to a slight refactoring since it will break the implemented waterfall model
it's just because; why would you light up Image Magick if it has nothing to do altogether; it would lead to additional "cost" in terms of processing, and most likely to image quality loss since it's piped through GM/IM unecessarily. Do you get what I try to say? :)
Ok, this makes sense. Will rework this once i have some time
Hi @veelenga,
thanks for your inspirational code that got me kickstarted to implement my own face detection and blurring.
While adapting, I realized that you could directly copy the "untouched" buffer to the final s3 bucket without directing it through GM, if there are no faces detected anyways.
Just wanted to suggest this.