Iterative deviation filter for fixed-valued impulse noise removal
- PDF / 1,319,076 Bytes
- 16 Pages / 439.37 x 666.142 pts Page_size
- 51 Downloads / 205 Views
Iterative deviation filter for fixed-valued impulse noise removal Jiayi Chen 1 & Yinwei Zhan 2 & Huiying Cao 1 Received: 30 March 2019 / Revised: 12 January 2020 / Accepted: 27 May 2020 # Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
An iterative deviation filter for fixed-valued impulse noise removal is proposed, with the aim to overcome the defects of existing filters, and further improve the denoising performance. In the proposed filter, a noise detection method based on the extreme intensity values and the deviation of neighbor pixels is proposed, i.e., the pixels with the extreme intensity and differ greatly from the mean of neighbor pixels, are identified as noises. A noise removal method based on the minimum deviation of neighbor pixels is proposed, i.e., the intensity of one neighbor noise free pixel, which is closest to the mean of neighbor noise free pixels, is used as estimated intensity of noisy pixel under consideration. Furthermore, the noise removal strategy performs iteratively and takes full advantage of the previous denoising results. Simulation results show that the proposed method has better denoising performance than the existing distinguished filters in terms of visual representation, peak signal to noise ratio and structural similarity index. Keywords Noise removal . Noise detection . Median filter . Correlation . Distribution . Deviation filter
* Yinwei Zhan [email protected] Jiayi Chen [email protected] Huiying Cao [email protected]
1
School of Biomedical Engineering, Guangdong Medical University, Zhanjiang 524023 Guangdong, China
2
School of Computer Science and Technology, Guangdong University of Technology, Guangzhou 510006, China
Multimedia Tools and Applications
1 Introduction Impulse noise is a frequent noise in image; it affects badly the image quality so that the image processing and analysis are severely affected. Thus, denoising is highly needed. There are two types of impulse noise, i.e., random-valued impulse noise and fixed-valued impulse noise. The latter is also called salt and pepper noise, corrupting image with extreme intensity values, i.e. 0 and 255; the noises with minimum intensity and the noises with maximum intensity distribute evenly and randomly with equal probability in corrupted image. For fixed-valued impulse noise removal, mean filters [3, 8, 11, 15, 20] and standard median filters [18] were initially proposed; however, the mean filters were found having no ability of detail and edge preservation because of its low-pass characteristic, and the standard median filters could not get a satisfactory result at high noise densities. Thereafter, as an improved version of standard median filter, the weighted median filter and centered weighted median filter [2, 12–14] were proposed; they replaces each pixel with weighted median of neighbor pixels. The drawback of the above filters lies in their undifferentiated treatment for the noisy pixels and the noise free ones, the noise free information are destroyed. In view of this, some re
Data Loading...
