Extracting a foreground actor from a video... [closed]
We have a series of video clips (single person speaking in front of a camera) and need to separate the actor from the background layer (to be composited onto a new layer with minimal artifacts). The good news is that camera is stationary, and in some scenes we even have shots of the background (mostly static) without the actor. I assume these two factors can be leveraged to make the process easier?
The question - What is currently the best way to do this? Does OpenCV have any built in libs we can use? The old photoshop + rotoscope method feels dated and time consuming... Surely there must be a better way!
Thanks!
what, exacty, using the opencv api, did you try ?
we -- that is, -- how many,exactly ?
We're a small team but we need to process LOTS of short video clips (as described above). We haven't dug in too deep into OpenCV because it is unclear if it can even help us. Our plan B is to just do things the old fashioned way (photoshop/aftereffects), but since that can be very time consuming, we're hoping OpenCV as some way to speed up the process.
*I'm assuming you asked about our team size because maybe you're aware of some expensive commercial software? If it does what we need it do, we would be willing to drop some cash (currently looking into Silhouette, but not sure how much better it is than PS/AE).
opencv is a library. you'd have to write your own program, using it. have a look at the tutorials
I am aware that OpenCV is a library. The question was, has anyone used this library to achieve such results?
I saw in the api that there is a BackgroundSubtractor, but are there any examples of it in action? Just how accurate is it?