The PCNN algorithm based on the visual perception of information can better analyze and understand the natural essence of image. But the multi-parameter settings and high computational complexity restrict the application of the algorithm in the industrial real-time image processing. Based on this, The parallel design and implementation of the PCNN algorithm based on the visual perception information is proposed. The PCNN algorithm is improved by the visual perception method, through analyzing parallelization of the feasibility of the algorithm, the parallel algorithm running at the cluster and experimental environment are developed. The performance of the parallel algorithm is verified by industrial image on the cluster, the experimental results show that the parallelized the PCNN algorithm has a better scalability and speedup.
