fswebcam

fswebcam is a good utility which works with Video4Linux version 1 and 2 devices. I've also used it with the so-called "driverless" USB Video class (UVC) devices. The main site for this utility is here. There is a problem associated with using fswebcam for stereo vision though, in that it's not intended to capture images from more than a single camera device at a time. As a workaround for this I modified the 20070108 version of fswebcam to be able to grab images from different devices in quick succession. The modified version, which includes all the original functionality, can be found here.

To install the modified version on an Ubuntu system first you will need to download the GD library upon which it is based. Move to the directory where you extracted the code, then go though the usual installation sequence:

./configure
make
sudo make install

Download the modified version of fswebcam and go through the same install procedure. You will now be able to use the utility with multiple devices, like this:

fswebcam -S 2 -d /dev/video0,/dev/video1 -r 320x240 --no-banner -save test.jpg

This will save two images called test0.jpg and test1.jpg. You can even capture from more than two camera devices if you wish, with the main limit being USB bandwidth.

Stereo vision / calibration GUI

Source code

Main project file: webcamstereo.mds

This is a Gtk based graphical user interface written in C# which can be used to test stereo vision algorithms and perform camera calibration, which was a spinoff development from the Surveyor stereo vision system. You can find more detailed instructions on calibration here.

The GUI may be cusomized by altering the settings within MainWindow.cs.

This program uses the modified version of fswebcam (see above) to grab images from camera devices. To prevent thrashing the hard disk with a lot of saving and loading of images you may wish to create a RAM disk, like so:

mkdir /home/myusername/ramdisk
sudo mount -t tmpfs none /home/myusername/ramdisk -o size=10m

A ten megabyte RAM disk should be more than adequate to store the temporary image files. You should then put the path which you created for the RAM disk into the temporary_files_path variable within MainWindow.cs.

Stereo vision server

Server source code

The stereo vision server developed for use with the Surveyor stereo camera can also be used with webcams, using the following syntax. As previously a RAM disk may be used to store temporary files.

stereoserver -leftdevice <left camera device, eg. /dev/video0>
             -rightdevice <right camera device, eg. /dev/video1>
             -broadcastport <port number on which to broadcast stereo feature data>
             -algorithm <"simple" or "dense">
             -calibration <calibration filename>
             -width <width of the image coming from the stereo camera in pixels>
             -height <height of the image coming from the stereo camera in pixels>
             -record <save raw and rectified images to the given path>
             -fps <ideal number of frames per second>
             -ramdisk <ram disk path where temporary files will be stored>

The client software also works identically, since it doesn't care what source the images came from.

stereoclient -server <IP address of the computer on which the stereo vision server is running>
             -port <broadcast port number on which the stereo vision server is communicating>