A Novel Parallel Approach to Image Processing for High-Performance Computing

Abstract As the utilization of big data and the computational requirements have increased the demand for superior image processing procedures has been of supreme importance. When it comes to the large amount of calculations that today’s high-resolution images and computationally heavy algorithms demand, serial computations do not suffice. This paper describes a new parallel algorithm for image processing suitable for application to high-performance computing. Taking advantage of the parallel computing environment and efficient algorithms which are used in the proposed approach, the response time is considerably improved. Data supporting this indeed can be best illustrated by the experiments conducted in this paper and they include filtering, edge detection and segmentation of images. The presented parallel strategy also minimized computational period and scaled effectively with enlargement and compounding of image collections. Such findings suggest significant enhancements in the performance outcomes in contrast to the results obtained from conventional methods suggesting the potential of parallel processing for bringing about revolutionary changes in image processing paradigms in high-performance computing environments.