Improv - Image Processing for Robot Vision
"See it Live at the Improv!"
Improv is a tool for basic real time image processing at
low resolution, e.g. suitable for mobile robots.
It has been developed for PCs with the Linux operating
system and runs under X Windows
Improv works with a number of inexpensive low-resolution
digital cameras (no framegrabber required).
Version 5.1 now on ftp server!
Improv 5 has been completely re-written, and now contains the
Improv also includes support for the following cameras:
- Modular design, allowing for easy integration of new functionality.
- Image processing operations are contained within plugins, compiled
linked files (.so), which can be loaded and unloaded
at run time.
- Improved support for still image sequences.
- An abstract Camera class, simplifying the creation of new camera drivers.
- A completely customisable interface, allowing for complex image processing sequences.
- Written using the QT Windowing library.
- Portable autoconf/automake build scripts.
Improv displays the source image (from either a still image
or a live camera) in the first window, while subsequent image
operations can be applied to this image in five more windows. For
each sub-window, a sequence of image processing routines may be
- Quickcam grayscale camera (parallel port) [no longer sold]
Make sure to set the protocol of your parallel port correctly
(boot-up setting to either ECP or EPP should work).
- Quickcam Color V2 camera (parallel port) [no longer sold]
Sorry, none of the newer QuickCams work - Connectix/Logitech
refuses to disclose the low-level interface.
- Various web cameras
with cpia driver, e.g. ZoomCam
- Video 4 Linux cameras
- and preliminary support for
Video 4 Linux 2 cameras.
Improv has been implemented by Michael Rudolph,
Thomas Bräunl, Elliot Nichols, Gerrit Heitsch, Uli Mezger,
Glen Horlin and Leon Koch
under direction of Thomas Bräunl.