Rights Creative Commons: Attribution 3.0 Hong Kong License Performance Analysis of Median Filtering on Meikom- a Distributed Multiprocessor System

This paper presents the performance analysis of real-izing median filtering on a distributed multiprocessor system. The results of the performance analysis give a good indication of the performance gain in using multi-processor for median filtering over uni-proces-SOT. Such performance gain is proportional to the problem size as shown by varying the size of the im-age. Furthermore, through the analysis, it is clear that the computation time and inter-processor com-munications scale well with the number of processors in the system. However, the overall system perform-ance does not have such behavior because of the ini-tialization overhead dominating the computation time as the number of processors increases beyond a cer-tain point. It is because of this relationship that an optimal performance is achievable with a certain number of processors. It is also found that this num-ber varies with the problem size. In nddition, the sub-image model is found to be an acceptable approach for this type of processing as only the necessary parts of the image are sent to the other processors. The master and slave scheme proves to be easy for pro-gramming, control and data manipulation. As a whole, this type of non-linear processing seem to fit well into the MIMD architecture.