Computer vision, owing to the size and complexity of its tasks and its
importance to industrial and economic growth, was selected as one of
the grand challenge problems by the U.S. Federal High Performance Comp
uting Program. Integration of vision operations is identified as a key
element of the challenge. A system to integrate computes vision in a
distributed environment is presented here. This system, called DeViouS
, is based on the client/server model and runs in a heterogeneous envi
ronment of Unix workstations. Modern computing environments include la
rge numbers of high-powered workstations connected by a very fast netw
ork. Many of these computers are idle most of the time. DeViouS takes
advantage of this feature of computing environments to distribute the
execution of vision tasks. Two primary goals of DeViouS are to provide
a practical distributed system and a research environment for vision
computing. DeViouS is based on a modular design that allows experiment
ation in various aspects of algorithm design, scheduling and network p
rogramming. It can make use of any existing computer vision packages w
ith very minor changes to DeViouS. DeViouS has been tested in an envir
onment of SUN and Digital workstations and has shown substantial impro
vements in speed over sequential computing with negligible overhead.