Parallel image processing for transputers based systems

Low and intermediate level image processing require a lot of computing power to be done in real time. Parallel processing seems to be the only economical way to achieve this power. The paper presents the development of parallel algorithms for image processing for MIMD distributed memory systems based on transputers. Improvements to the sequential image processing techniques are proposed, in order to make the parallel algorithms more suitable for this type of architectures. The parallel versions of the image processing algorithms have all been integrated in a modular software package which provides a simple and user friendly interface. Evaluation of the performance of the multitransputer systems is also made, and the results obtained can be useful for those which are involved in the design of parallel image processing systems