koobas.hobune.stream
computer vision (openCV) to create / control 2D - colliders....?
Hi there,
I am trying to make a game where video-input origins / create / controls 2D-colliders.
The "case" is that I turn video input to greyscale and the "dark" parts of video create / format 2D-collider and the "light" parts of video is free-space where you can play.
I am trying to figure out how can this be possible and I think that the answer in in simple computer vision. Looking around the internet I have find some "glimpses" of solutions, in open-CV, but nothing clear whether this can happened with a certain stability or it is just hypothesis..
In fact, I am not sure what I am looking for.. Lets say that I have a black 'n' white (2-colour) video. I What am I looking for?...:
A way to turn black to a 2D-collider?
a way to turn black to a 2D -sprite and then to a collider?
Get spline points of black shapes and turn them to polygon collider?
...........?
And how...
Can anyone help me? Moreover than a simple yes/no to CV, any ideas for any kind of solutions are welcome. Thanx...
