Hi
For a project I’m working on, I need to render multiple pedestrians per image.
After rendering the image, I want to calculate the potential overlap between pedestrians. For example, if a pedestrian is occluded more than X percent by another pedestrian, I need to know this.
I think the easiest way to do this, would be to get a separate segmentation map for every pedestrian. Then I can easily check the overlap with a tool like Matlab.
I know you can create segmentation maps for all pedestrians/objects at the same time by rendering the object ids as colours.
But is it also possible to render e.g. the object ids for every object separately in an efficient way.
I could make every pedestrian visible 1 by 1 and then render, but this would take quite long since I need to simulate millions of images.
So, can someone help me finding an efficient way to render those segmentation maps separately?
Thank you very much!
