VideoMan is an open-source C++ library that helps you developing video based applications. We created VideoMan to increase our productivity developing computer vision applications, but it can be used in many other ways.
With VideoMan is very easy and fast to initialize and acquire images from any kind of video input like cameras(webcameras, USB, firewire, industrial cameras, GigE,...), video files, frame grabbers, 3D sensors (Kinect) and image sequences. You can initialize multiple inputs at the same time and show them in the screen. Also, you can show the results of the processing steps of your computer vision algorithm (extracted edges, back projections, detected blobs...) This way you can see the results of what are you coding, checking the results of your algorithm steps. Moreover, the implemented visualization scheme make easier to combine computer graphics with video, for example to show a 3d model on top of video for augmented reality applications.
It is very easy to use VideoMan as a camera input library for OpenCV extending the Highgui functionality.