In this thesis, 3D range image acquisition and reconstruction procedures are implemented in a computer, called FACS(Factory Automation Computer System) and a robot arm is manipulated using its data. The major features of FACS include the real time processing capability supported by real time operating system, the communication capability supported by a token bus networks, the powerful computing performance supported by Transputers and the user interface supported by UNIX system V. The experimental environment for this thesis work is consisted of two FACS computer nodes. One node acquires images using camera and frame grabber, processes them using Transputers, calculates the position of an object and transfers it via token bus communication network. The other node receives it and manipulates a robot arm, Rhino, to point to the wanted object. It takes nearly 125 msec for the range image acquisition and its reconstruction. It is also proven that the communication and the robot manipulation modules of the FACS operate well. As an application to 2D image processing using parallel processing module of FACS, image convolution is performed with 1,4,8,16 processors in the parallel module. It shows that the speed-up factor is linearly proportional to the number of processors in this case.
