Parallelisation of the overlap-add block-filter algorithm for image processing

The utilization of multiple DSP systems for parallel processing in computationally demanding applications is growing fast. The field of 2-D imaging and multidimensional signal processing is a good candidate for parallel processing due to the large amount of low level computations required. In this paper we present a 2-D parallel algorithm for the overlap-add digital block filter and a multiprocessor architecture suitable for image processing applications. We also investigate parallel processing systems in terms of performance and efficiency. The implementation of the 2-D parallel algorithm is discussed and results are given to show the performance of the parallel system compared to a single processor system.