Gaussian Noise Filter Using Variable Step Time Matrix of PCNN

Abstract:

A novel method was brought forward for the purpose of filtering Gaussian noise effectively by using variable step time matrix of the simplified pulse coupled neural network (PCNN). Firstly, the time matrix of PCNN, related to the grayscale and spatial information of an image, is calculated to identify the noise polluted pixels. Subsequently, a variable step, a long step for strong noise and a short step for weak noise, based on the time matrix is applied to modify the grayscale of noised pixels in a sliding window. And then wiener filter is used to the image to further filter the noise. Experiments show that the proposed filter can remove Gaussian noise effectively than other noise reduction methods such as median filter, mean filter, wiener filter etc, and the filtered image is smooth and the details and edges are sharp. Compared with existing PCNN based Gaussian noise filter, the proposed filter gets higher Peak Signal-to-Noise Ratio (PSNR) and better performance.

• Periodical: Applied Mechanics and Materials (Volumes 48-49)
• Pages: 551-554
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.48-49.551
• Online since: February 2011
• Authors: Yuan Yuan Cheng, Hai Yan Li, Qi Xiao, Yu Feng Zhang, Xin Ling Shi
• Keywords: Gaussian Noise, Peak Signal-to-Noise Ratio (PSNR), Simplified Pulse Coupled Neural Network (PCNN), Variable Step Time Matrix

References

[1] Ko S J, Lee Y H. in: Center weighted median filter and their application to image enhancement [J]. IEEE Trans. On cir. Syst., 1991, 38: 984~993.

DOI: 10.1109/31.83870

[2] Ma Yide, Shi fei, Li Lian., in: Gaussian noise filter based on PCNN[C]. IEEE ICNNSP, 2003, 1(1): 149~151.

[3] Shi M, Zhang J, Zhu X. in: A Method of Image Gaussian Noise Filtering Based on PCNN (in Chinese) [J]. Computer Applications, 2002, 22(6): 1~4.

[4] Shi M, Mao J, Liang Y., in: Method for filtering image contaminated with strong Gaussian noise (in Chinese) [J]. Computer Applications, 2007, 27(7): 1637~4.

[5] Ma Y, Lin D, Zhang B, Liu Qing etc, in: A novel algorithm of image Gaussian noise filtering based on PCNN time matrix[C]. 2007 IEEE International conference on signal processing communication, 2007: 1499~1502.

DOI: 10.1109/icspc.2007.4728615

[6] Wang Z, Ma Yid, Cheng F etc, in: Review of pulse-coupled neural networks [J]. Image and computing, 2010, 28: 5~13.