Long shot question about OpenCV

[Mark Wieder]

On 10/17/22 15:00, Paul Dupuis via use-livecode wrote:
> Has anyone out there done any work at wrapping the OpenCV computer 
> vision/image recognition open source library with LCB?
> See https://opencv.org/
> 
> I am particularly interested in simple face and/or object recognition in 
> images - something where I can pass the library an image file (like a 
> PNG) and it can return whether or not there is a face in the image and 
> if yes/true, the rectangle within the image that bounds what it thinks 
> is a face (or faces as I suppose it could return multiple rects).

Haven't tried this with LC (only done OpenCV with python), but for image 
recognition tensorflow is pretty easy to leverage with javascript. Takes 
a bit of training though. Might want to start with a pretrained model:

https://basilebong.com/blog/image-recognition-with-tensorflow-js-1m74

-- 
  Mark Wieder
  ahsoftware at gmail.com