文献类型：期刊文献

英文题名：Adaptive Noise Detector and Partition Filter for Removing Impulse Noise from Grayscale Images

作者：Lin, Cong[1,3]; Qiu, Chenghao[1]; Xiao, Xiuchun[3]; Feng, Siling[1]; Feng, Mengxing[1,2]

机构：[1] College of Information and Communication Engineering, Hainan University, Haikou, 570228, China; [2] State Key Laboratory of Marine Resource Utilization in South China Sea, Hainan University, Haikou, 570228, China; [3] College of Electronic and Information Engineering, Guangdong Ocean University, Zhanjiang, 524088, China

年份：2022

外文期刊名：SSRN

收录：EI(收录号：20220103773)

语种：英文

外文关键词：Adaptive filtering - Image denoising - Iterative methods - Medical imaging - Noise abatement - Pixels

外文摘要：The random-value impulse noise (RVIN) denoising method based on the preset detection threshold or local window information does not have good generalization performance and edge-preserving denoising effect. In this paper, we present a novel two-stage method to remove RVIN. Based on the idea of pixel clustering and grouping, we divide all pixels in the damaged image into several categories according to the characteristics of gray distance similarity, and then iteratively solve the optimal detection threshold of each group to identify the noise. In the noise removal stage, we propose a partition decision filter. For the noise pixels in flat and detail areas, LCI weighted filter and edge direction filter are designed respectively to recover the pixels damaged by RVIN. The experimental results show that the proposed method not only has good noise reduction performance and generalization performance for both natural and medical images with medium and high noise levels, but also outperforms other advanced filtering techniques in terms of visual effects and objective quality evaluation. ? 2022, The Authors. All rights reserved.