Improvement of a Parallel System for Image Processing

Abstract

Digital images are digital signals captured through different means. Sometimes these captured images contain variations combined with the original signal, this variations are called noise. Echo is a particular kind of noise with characteristics that turns it into a very interesting problem to solve. The echo detection process and its subsequent elimination from a digital Image involves extensive mathematical calculations. Differents Parallel approaches taking advantage of new architectures can be implemented to solve this problem. Nevertheless these approaches have time depending characteristics, so the Processing time is still the critical point. One improved version of a Parallel one may be implemented by using a different algorithm and some other techniques that much more reduce the Processing time. In this work, the authors discuss their earlier work, the present approach and the future directions of this experimental application. Finally the resulting values are sketched.