Open Access

Research on the Image Denoising Method Based on Partial Differential Equations

   | Oct 20, 2016
Cybernetics and Information Technologies's Cover Image
Cybernetics and Information Technologies
Issue Title: Special Issue on Application of Advanced Computing and Simulation in Information Systems


In this paper we propose a new approach for image denoising based on the combination of PM model, isotropic diffusion model, and TV model. To emphasize the superiority of the proposed model, we have used the Structural Similarity Index Measure (SSIM) and Peak Signal to Noise Ratio (PSNR) as the subjective criterion. Numerical experiments with different images show that our algorithm has the highest PSNR and SS1M, as well as the best visual quality among the six algorithms. Experimental results confirm the high performance of the proposed model compared with some well-known algorithms. In a word, the new model outperforms the mentioned three well known algorithms in reducing the Gibbs-type artifacts, edges blurring, and the block effect, simultaneously.

Publication timeframe:
4 times per year
Journal Subjects:
Computer Sciences, Information Technology