期刊名称：International Journal of Signal Processing, Image Processing and Pattern Recognition
印刷版ISSN：2005-4254
出版年度：2015
卷号：8
期号：8
页码：273-282
DOI：10.14257/ijsip.2015.8.8.29
出版社：SERSC
摘要：In recent years, the theory of partial differential equations (PDE) for its rigorous mathematical theory foundation has been widely used in the fields of image processing. Medical imaging seeks to reveal internal structures hidden by the skin and bones, as well as to diagnose and treat disease. In order to detect the noise in the medical images, many models are studied, but the noises in medical images are much more complex than typical images. This paper introduces a new image noise detection approach using morphology and partial differential equations where are based on the morphology reconstruction with anisotropic diffusion to make full use of the advantage of Catte model. This proposed approach has been tested with the biomedical cell images with comparing with the Catte model, PM model and Canny model. The experimental results show that this proposed approach outperforms the other three models in terms of defined indicator and efficiency
