Topics

1

Face and Body Scanning / Improved masking options for human scanning

« on: April 03, 2014, 08:29:34 PM »

Hi all.  We've been using PhotoScan's built-in tool to get masks from background images (i.e. a second set of clean plate/empty background images taken right after the subject is photographed).  It does quite a good job, but seems to have the following issues:

1) the edge of the mask is often too generous in letting in the background, which lets through the white background; this is particularly problematic when creating textures for hair and areas where 'webbing' is likely (fingers, armpit, crotch)
2) it often ends up masking out dark portions of the subject if they are in the same spot as a camera lens in the background image
3) due to diffuse reflection of the subject on our white floor, the contact point of their feet with often includes a significant portion of the floor around their feet

So far we have been manually cleaning up these problem features using the PhotoScan tools, which ends up taking a good 60-90 seconds per photograph on average.  Obviously I'd like to take this number down to zero, so have been looking at alternative methods for background subtraction, which I could then feed into PhotoScan.  This project looked like a good option, but ended up having its own problems, and being rather slow:

http://docs.opencv.org/trunk/doc/tutorials/video/background_subtraction/background_subtraction.html

Is anyone else using external programs for mask generation?  I should know the answer to this but for some reason it is eluding me...

Thanks!

2

Face and Body Scanning / Max cameras per computer over USB

« on: January 16, 2014, 06:56:02 AM »

Hi all, coming up against the USB max devices problem here in building out our array.

We have been using 12 cameras connected to a single Windows 8 machine, works great.  However upon plugging in a 13th camera, Windows 8 rejects it, citing "Not Enough USB Controller Resources".  I suspect that each camera is getting allocated a certain amount of USB bandwidth, and because the cameras only connect via USB 2.0, they run out of the needed bandwidth rather quickly.

To compensate, I got a PCI-e USB 3.0 controller card, hoping that a dedicated card might fare better, or at least allow another 12 to be connected.  However the card only allowed 7 more cameras to be connected before throwing up the same error.  Which seems bizarre, since the PCIe x1 slot that it's in should have adequate bandwidth available to it (500Mbps) that USB 2.0 requires (480Mbps)...could it be the B85 chipset on the motherboard coming up short somehow?

http://www.newegg.com/Product/Product.aspx?Item=17Z-0002-00002

Unfortunately I am unable to conclusively confirm the exact problem on my Windows 8 machine, as I cannot find a utility for querying each device's bandwidth allocation (it used to be conveniently available in Win7's Device Manager).  Does anyone have a favorite utility for checking this in Windows 8?

The above aside, is everyone just going the 12 cameras per computer route, using minimal machines, and having them all write to a shared network location?

3

General / Test Renders - Camera Calibration Settings

« on: January 14, 2014, 11:37:05 AM »

As suggested by others on these awesome forums, we have been working on Maya-based renders to test camera positions and coverage, lens lengths, and use of noise projection.  There are two aspects that are unclear though, regarding camera calibration:

1) What is the pixel size for these renders, since there is no real sensor?

2) Because Maya's renders are 'perfect', there is no lens distortion to undistort, and the diminishing effects of wider lenses like 24mm cannot be accurately simulated.  Does anyone have a Photoshop or Nuke script that can recreate the lens distortion using the same Brownian model that PS uses to remove it?

Any other tips for making accurate simulations?

Thanks!