Sequentially combined mean-median filter for high density salt and pepper noise removal

Detecting Breast Cancer (BC) at an early stage can reduce the rate of mortality. A mammogram is a preliminary examination of breast cancer with an optimistic and accurate method, for the same test will be used to know the behavior of the tumor in all stages of cancer. Unpolished Images need to process for a clear image to analyze, and enhanced images can use to view the tissue, muscle, and tumor. Image processing methods are not only used to remove the noisy pixels but also help to extract the Region of Interest (RoI) from the image. The size and structure of the tumor will be extracted using statistical analysis. This study was experimented on mammogram images by analyzing the types of noise on the image and filtering methods applied for a specific class of noise data. The different filtering approaches like Adaptive, Gaussian, Threshold, Mean, Mean-Median, Bilateral, Wiener, and Tri-State Filters were used. Mammogram images are grayscale images. The objective of this research experiment is noise data will reduce the detection factors of the tumor with True−Ve and False+Ve results. Fault results will impact the accuracy to identify the RoI. Image Preprocessing will help to improve the quality of the images and help in the treatment plan and expert advice in the decision of the disease stage. Results: The Dataset used images from CBIS-DDSM (The Curated Breast Imaging Subset of DDSM-[Digital Database for Screening Mammography]). The obtained Peak-signal-to-Noise Ratio (PSNR) achieved is 80% minimum.

show abstract

“…The overall advantage of the mean filter is single noise pixel will not impact the view of the image. 12 f(x, y) = 1 mn…”

Section: Mean Filtermentioning

confidence: 99%

“…Obtained results are filtering methods on images Mean Square Error has a drawback like an image intensity scaling, the metric used to measure PSNR values are dB. Pixel values are defined as S, the optimal values are PSNR is 30 dB 12. …”

mentioning

confidence: 99%

Removal of noise on mammogram breast images using filtering methods

Mudrakola

Hegde²

2022

Concurrency and Computation

View full text Add to dashboard Cite

show abstract

“…This introduced an artifact in the enhanced images as the neighborhood details were not efficiently maintained. Later a Sequentially Combined Mean Median Filter (SCMMF) [12] was proposed which efficiently reduced salt and pepper noise for some high noise densities. This algorithm however failed to properly enhance images with noise densities greater than 80%.…”

Section: Related Workmentioning

confidence: 99%

Impulse Noise Removal Algorithm Using Modified Adaptive Distance-Related Weighted Mean Filter

Halder¹,

Choudhuri²

2021

Advances in Intelligent Systems and Computing

View full text Add to dashboard Cite

Reduction of impulse noise is a common preprocessing step for many image processing and computer vision applications. Salt-and-pepper noise is a common type of impulse noise which is obtained during image acquisition, storage, and transmission. In such a scenario, the image histogram has a high peak at both ends, leading to a corrupt image caused by the noisy pixels having maximum or minimum intensity values. In the proposed method, an adaptive distance-related weighted mean filter has been implemented to remove impulse noise from a corrupt image. In this technique, each noisy pixel is restored by the distance-related weighted mean of the first occurring non-noisy pixel in every immediate neighboring direction and thus it improves the quality of the image. The proposed technique is tested on standard sample images and it produces better quantitative results than the other existing state-of-the-art algorithms.

show abstract

“…These sudden intensity changes not only degrade the images but also hamper the successive image processing procedures like morphological processing, segmentation, object recognition etc. Impulse noise is of two types: salt and pepper noise (SAPN) [2] and random valued impulse noise (RVIN) [3]. Both the types of noises disturb the homogeneity of the pixel"s intensities of images.…”

Section: Introductionmentioning

confidence: 99%