Conventional Background Removal involves checking pixel by pixel differences in the background image and the foreground image. This involves 2 required images which are taken at the same spot. A small variation will fail the algorithm.
We are trying to remove background using only 1 image. But we require the user to pose in such a way that our algorithm to work, and we will need the user to be in front of a plan background, preferably in front of a flat colour background.
We are still currently testing its performance but initial performance aren't very good.
Subscribe to:
Post Comments (Atom)
Posts
No comments:
Post a Comment