中文摘要：

利用PVM建立PC机群环境，并在此环境中开展并行研究。分别采用串行与并行的计算方法实现了图像的平滑处理。在并行计算中。对大小不同的图像分别进行了并行计算，得出对应的处理时间以及所获得的加速比。最后采用了一种改进的算法实现了图像处理。针对所获得的实验数据，得出加速比较低，通过采用数据本地化的优化措施．有效地提高了并行加速比．达到了良好的性能效果。通过对实验结果的分析可知，当前的并行处理的实施对于处理数据传输量很小的求解问题上有一定的性能效果．只有网络性能达到一定程度的时候，并行处理就可以应用在普通问题规模上．并能够取得良好的处理速度。

英文摘要：

Parallel computing are analyled in the PC Cluster that based on PVM. In this paper, serial algorithm and parallel algorithm has been used in image smoothing processing. In parallel computing process, the two images with different sizes have gotten different times and different speed-up ratio. In order to improve the performance, data localization is proposed to conquer the low speed-ratio. The results show that the speed-up ratio is restricted by the performance of network. If the network performanced well,the image processing will win the better result.