Abstract: A fast image filtering algorithm based on Perona and Mailik’s anisotropic diffusion model is put forward for solving the problem that the requirement for specialty paper defects detecting can not be met in the existing methods. Perona and Mailik’s anisotropic diffusion model is adopted to protect those subtle edges of specialty paper defects while images for paper being filtered. With the help of SMP system and multi-threading technology, Perona and Mailik’s anisotropic diffusion is performed by the parallel iterative algorithm to ensure the rapidity of image filtering algorithm for specialty paper defects detection. Experimental result shows that adopting this method, subtle edges of specialty paper defects are able to be detected and the real-time performance of this method is significantly better than the traditional implementation of Perona and Mailik’s anisotropic diffusion model filtering.

Key words: specialty paper defects, Perona and Mailik’s anisotropic diffusion, parallel iterative diffusion, Symmetrical Multi-Processing（SMP） system, multi-threading technology

摘要： 为解决现有纸病检测方法中，采用的滤波算法性能无法达到部分特种纸的纸病检测要求的问题，提出一种基于P-M方程扩散模型的快速图像滤波方法。通过采用P-M方程多次迭代扩散进行滤波，滤波同时保护特种纸纸病的细微边缘；并借助SMP系统以及多线程技术对P-M方程扩散过程实现并行迭代算法以保证整个纸病检测算法过程的快速性。实测结果表明，采用该方法，系统可以有效检测到特种纸上表征微弱的纸病，并较传统P-M滤波方法的实现在实时性能上有了大幅提高。

关键词: 特种纸纸病, P-M方程, 并行迭代扩散, 对称多处理（SMP）系统, 多线程技术