Just started using your program and have been enjoying it. Well done, and thanks for putting in the time.
I'm having some trouble getting the depth extraction to work the way I want it to in Intrael. Either I'm doing something wrong, or I need to change my expectations. My understanding from the latest docs is that a value of e=0 does no background extraction (intrael detects anything between z and Z as an object), and both e > 0 and e < 0 generate a reference frame (whatever Intrael sees as present in the viewing area when it builds the frame) and both do different things with it. My desired functionality is pointing the kinect at a given room, running background extraction, and then having Intrael only report objects that are closer to the kinect than the background reference frame (displaying them as white/gray and including them in the blob detection), while ignoring the background reference frame entirely (displaying as black and being excluded from blob detection).
What I'm seeing is that e > 0 extraction seems to recognize the background, but still display it as present (white/gray) and detect it as an object blob and display any objects that move between it and the kinect as an absence of space (black). This causes the background to be one big object blob split by anything that moves between it an the kinect. This is exactly the inverse of what I want.
I expected e < 0 extraction to do the opposite of e > 0 extraction, and assumed it's what I want. However, when I try to use it, I only see alternating gray and black vertical bars and no video or object detection. Should e < 0 do what I'm looking for? If so, what do I need to change to make it work properly?
I'm running an AMD x86 processor on Ubuntu 14.10 with the latest code and dependencies (as of last week). According to the Ubuntu system monitor, the program is not resource locked in any way.
Any help or advice would be appreciated. Thanks!