一种基于多尺度噪声检测的图像中值滤波器

介绍了标准中值滤波与有效中值滤波的概念,提出了一种基于自适应多尺度噪声检测的中值滤波器,可用于恢复被椒盐脉冲噪声污染了的图像。滤波器将输入图像像素分为有效信号类、脉冲噪声类和恒定区域类,对各类像素采用不同的方法进行滤波处理。实验结果证明,本文算法的性能比现存的其它许多算法有了显著的提高,而且便于实现。     

两阶段多层感知的随机脉冲噪声比例预测

目的 基于卷积神经网络（CNN）在图块级上实现的随机脉冲噪声（RVIN）降噪算法在执行效率方面较经典的逐像素点开关型降噪算法有显著优势,但降噪效果如何取决于能否对降噪图像受噪声干扰程度（噪声比例值）进行准确估计。为此,提出一种基于多层感知网络的两阶段噪声比例预测算法,达到自适应调用CNN预训练降噪模型获得最佳去噪效果的目的。方法 首先,对大量无噪声图像添加不同噪声比例的RVIN噪声构成噪声图像集合;其次,基于视觉码本（visual codebook）采用软分配（soft-assignment）编码法提取并筛选若干能反映噪声图像受随机脉冲噪声干扰程度的特征值构成特征矢量;再次,将从噪声图像上提取的特征矢量及对应的噪声比例分别作为多层感知网络的输入和输出训练噪声比例预测模型,实现从特征矢量到噪声比例值的映射（预测）;最后,采用粗精相结合的两阶段实现策略进一步提高RVIN噪声比例的预测准确性。结果 针对不同RVIN噪声比例的失真图像,从预测准确性、实际降噪效果和执行效率3个方面验证提出算法的性能和实用性。实验数据表明,本文算法在大多数噪声比例下的预测误差小于2%,降噪效果（PSNR指标）较其他主流降噪算法高24 dB,处理一幅大小为512×512像素的图像仅需3 s左右。结论 本文提出的RVIN噪声比例预测算法在各个噪声比例下具有鲁棒的预测准确性,在降噪效果和执行效率两个方面较经典的开关型RVIN降噪算法有显著提升,更具实用价值。     

Self-adaptive algorithm of impulsive noise reduction in color images

In this paper a new approach to the problem of impulsive noise reduction in color images is presented. The basic idea behind the new image filtering technique is the maximization of the similarities between pixels in a predefined filtering window. The improvement introduced to this technique lies in the adaptive establishing of parameters of the similarity function and causes that the new filter adapts itself to the fraction of corrupted image pixels. The new method preserves edges, corners and fine image details, is relatively fast and easy to implement. The results show that the proposed method outperforms most of the basic algorithms for the reduction of impulsive noise in color images.     

去除椒盐噪声的自适应开关加权均值滤波

提出一种去除椒盐噪声的自适应开关加权均值滤波算法。该算法采用一种新的噪声检测方法将图像中的像素分为信号点和噪声点,对检测出的噪声点采用加权均值滤波进行处理,而信号点保持其灰度值不变直接输出。实验结果表明,该算法能在有效去除椒盐噪声的同时保护图像细节,较传统中值滤波及其改算法有更好的滤波性能。     

一种新的高密度椒盐噪声滤波算法

针对现有算法对高密度椒盐噪声滤波不足的问题,提出一种新的高密度椒盐噪声滤波算法。通过噪声检测将含噪图像的像素分为信号点和噪声点,对每一个椒盐像素,计算以该像素为中心的窗口内非椒盐像素中值,信号点则保持其灰度值不变直接输出,重复以上过程,直到没有噪声点被替换。实验结果表明,该算法能在有效去除椒盐噪声的同时保护图像细节,较传统中值滤波及其改进算法有更好的滤波性能。     

Restoration of images corrupted by Gaussian and uniform impulsive noise

Many approaches to image restoration are aimed at removing either Gaussian or uniform impulsive noise. This is because both types of degradation processes are distinct in nature, and hence they are easier to manage when considered separately. Nevertheless, it is possible to find them operating on the same image, which produces a hard damage. This happens when an image, already contaminated by Gaussian noise in the image acquisition procedure, undergoes impulsive corruption during its digital transmission. Here we propose a principled method to remove both types of noise. It is based on a Bayesian classification of the input pixels, which is combined with the kernel regression framework.     

基于噪声检测的遥感图像模糊滤波

提出了一种基于噪声检测的遥感图像模糊滤波方法。该方法首先用一个噪声检测标准将输入图像的像元分为为噪声像元和信号像元,然后,利用模糊数学的相关理论对噪声像元进行处理,并把处理结果赋给输出图像的对应像元;而对于信号像元,则不进行处理,直接把它的值赋给输出图像的对立像元。另外,在图像处理过程中,把输入图像的噪声像元用其处理结果代替,以更好地改善图像滤波处理结果。实验结果表明本文的方法能有效地去除图像中的椒盐噪声。     

基于二级修复的多方向加权均值滤波算法

针对目前算法不能有效去除高概率的椒盐噪声并保护图像边缘和细节特征的缺点,提出了一种基于二级修复的多方向加权均值滤波算法。在噪声检测阶段,首先利用一个方差参数判断当前像素点与其邻域像素点之间的灰度差异程度,再通过将方差参数和灰度极值相结合的方法检测出图像中的椒盐噪声点。在噪声修复阶段,提出一种二级修复方法来修复噪声点的灰度值。首先利用改进的自适应中值滤波器对椒盐噪声点进行第一级噪声修复;然后利用方差参数将第一级修复后的噪声点划分为两类,并采用不同的修复方法对这两类像素点进行第二级噪声修复,一类像素点采用均值滤波器进行再修复,另外一类像素点采用多方向加权均值滤波器进行再修复。数值实验结果表明,所提算法的滤波性能和边缘保护能力均优于当下很多先进的滤波器。     

Robust edge detection

Edge detection is an important issue in computer vision and image understanding systems. Most conventional techniques have assumed Gaussian noise, and their performance could decrease with the departure of noise distribution from normality. In this paper, we present an edge detection approach using robust statistics. The edge structure is first detected by a robust one-way design model, and then localized by a robust contrast test. Finally, hysteresis thresholding is applied to yield the output edge map. To evaluate its performance, experiments were carried out on synthetic and real images corrupted with both Gaussian noise and a mixture of Gaussian and impulsive noise. The results show that the performance of the proposed edge detector is stable and reliable under severe impulsive noise conditions.     

Assessment of potential noise-induced hearing loss with commercial “Karaoke” noise

A study was conducted to assess the noise levels of popular Karaoke environments in Korea and the degree of potential noise-induced hearing loss due to commercial Karaoke noise. Using 18 subjects with normal hearing, a two-way, mixed-factors experimental design was used with two independent variables of “noise source” (no-singer, one-singer, and two-singer conditions) and “music type” (Trot, Ballad, and Rock music). For each experimental condition, average sound pressure levels and maximum sound pressure levels were measured. For assessing amount of temporary hearing threshold shift as a measure of potential hearing loss index, pure-tone audiometry was applied for measuring subjects’ hearing threshold levels of both ears before and after 100 min exposure to Karaoke noise. Questionnaires from 155 actual Karaoke users were also obtained to evaluate realistic user subjective perception on the Karaoke environment. Results showed that noise levels of typical Karaoke singing environments were higher than 95 dBA, with maximum noise levels often exceeded the US OSHA's non-permissible 115 dBA level. Further statistical analysis of hearing threshold shift revealed that up to 8 dB of significant hearing loss was found at the most important human hearing frequency band, centered at 4000 Hz, after about less than 2 h of Karaoke noise exposure, indicating that Karaoke facilities may pose a serious threat to noise-induced hearing loss. Along with some ergonomic/safety issues, practical hearing protection strategies are suggested and discussed.

Relevance to industry

Since noise levels from popular commercial Karaoke facilities are found to be potentially dangerous, the Karaoke-related industries, of supporting a big consumer market of entertainment, need to provide safer environments to protect naive users from potential hearing loss. Providing better design of Karaoke facilities with some ergonomic intervention strategies (such as proper sound absorptive treatment in Karaoke rooms, displaying warning signs of potential hearing loss, and setting up an upper safety volume limit in the Karaoke machine, etc.) may help those industries not only contribute for consumer protection but also develop better market with strong ethical and legal support in the future.     

分块策略实现图像椒盐噪声密度估计

目的 椒盐噪声是造成图像污染的常见因素之一,椒盐噪声密度的估计对椒盐去噪过程中滤波窗口大小的选择具有指导作用。为此提出了一种基于分块策略的椒盐噪声密度估计算法。方法 首先对图像按行列等分后形成多个图像子块,统计每个子块中灰度为0或255的像素点个数并排序,然后根据排序后个数差分值函数特征对子块进行筛选,最后将所有候选子块噪声密度估计值的中值作为对整幅图像噪声密度的估计。结果 为验证算法的有效性,选取了两组不同类型的图像进行仿真,与现有椒盐噪声密度估计算法对比噪声密度估计结果。仿真实验结果表明,当图像自身包含较多灰度为0或255的像素点时,本文算法的噪声密度估计精度优于现有各种算法,标准差比现有算法小近一个数量级。当图像自身不包含灰度为0或255的像素点时,本文算法也能达到现有算法中最优的估计效果。结论 本文算法不仅能准确估计不同强度下的噪声密度,而且适用于自身包含灰度为0或255的像素点多的椒盐噪声图像。     

2-D DOAs estimation in impulsive noise environments using joint diagonalization fractional lower-order spatio-temporal matrices

A novel joint diagonalization fractional lower-order spatio-temporal （ST） moments DOA matrix method is proposed to estimate the 2-D DOAs of uncorrelated narrowband signals in the presence of impulsive noise. The new method retains the advantage of the original ST-DOA matrix method which can estimate 2-D DOAs with neither peak searching nor pair matching. Moreover, it can handle sources with common 1-D angles. Simulation results show that the proposed method yields to better performance to restrain the strong impulsive noise than ST-DOA matrix method, especially for low signal-to-noise ratio case.     

Fast averaging peer group filter for the impulsive noise removal in color images

In the paper, a new approach to the impulsive noise removal in color images is presented. The new filtering design is based on the peer group concept, which determines the membership of a central pixel of the filtering window to its local neighborhood, in terms of the number of close pixels. Two pixels are declared as close if their distance in a given color space does not exceed a predefined threshold value. A pixel is treated as not corrupted by the impulsive noise process, if its peer group consists of at least two close pixels, otherwise this pixel is replaced by a weighted average of uncorrupted samples from the local neighborhood. The peer group size assigned to each pixel is used for the averaging operation, so that pixels which have many peers are taken with higher weight. The new filtering design proved to restore efficiently color images corrupted by even strong impulsive noise, while preserving tiny image details. The beneficial property of the proposed filter is its very low computational complexity, which allows its application in real-time image processing tasks.     

基于CIM的偏差补偿稀疏NLMAD算法研究

为了解决输入信号受噪声干扰和输出观测噪声具有脉冲特征的稀疏系统辨识问题,提出一种基于CIM的偏差补偿NLMAD(Normalized least mean absolute deviation, NLMAD)算法。 利用NLMAD算法可有效抵御脉冲输出观测噪声的优势,首先应用无偏准则设计偏差补偿NLMAD算法来有效解决由于输入噪声导致的估计偏差问题。再次考虑到稀疏系统辨识问题,将CIM作为稀疏约束惩罚项引入到偏差补偿NLMAD算法提出了新的稀疏自适应滤波算法CIMBCNLMAD。将所提算法应用于输入和输出均含有噪声的稀疏系统辨识和回声干扰抵消场景中,实验表明CIMBCNLMAD算法的稳态性能优于其它自适应滤波算法,说明该方法具有强的鲁棒性且可应用于工程实践。     

A “mixed” small gain and passivity theorem in the frequency domain

We show that the negative feedback interconnection of two causal, stable, linear time-invariant systems, with a “mixed” small gain and passivity property, is guaranteed to be finite-gain stable. This “mixed” small gain and passivity property refers to the characteristic that, at a particular frequency, systems in the feedback interconnection are either both “input and output strictly passive”; or both have “gain less than one”; or are both “input and output strictly passive” and simultaneously both have “gain less than one”. The “mixed” small gain and passivity property is described mathematically using the notion of dissipativity of systems, and finite-gain stability of the interconnection is proven via a stability result for dissipative interconnected systems.     

基于排序跳变点的脉冲噪声检测与滤除算法

提出了一种简单的判断脉冲噪声点的算法。该算法先将滤波窗口内的像素值按照大小排序,然后找到排序后像素值的两个最大跳变点,如果此跳变点发生在靠近最大值或最小值处,并且跳变幅度超过文中实验得出的域值,就可以根据窗口中心像素值的位置,即它在两个跳变点所确定的范围之内还是之外,将其判为信号点或噪声点,最后用判断出的信号点对噪声点进行恢复。实验结果表明,该算法能够几乎准确地判断出噪声点与信号点,从而达到了更好的去噪效果。     

基于模糊推理的多窗口中央值滤波器设计方法

提出了一种基于模糊推理用于去除图像椒盐噪声的中央值滤波器的新型设计方法,在图像复原处理中,理想的期望是对图像被劣化的部分处理,没有被劣化的部分不作处理,但实际图像处理中处理点是否为噪声点具有模糊性.利用模糊推理对处理点像素多大程度上属于劣质像素进行推定,并且多个模糊滤波器联合使用,处理结果证明对广范围噪声发生率的各种被椒盐噪声劣化的图像复原处理都适用.     

消除椒盐噪声的改进滤波算法

数字图像在采集、传输等过程中会产生椒盐噪声。传统滤波算法在高噪声率情况下,很难对图像进行有效处理。该文在极值中值滤波的基础上,提出一种具有精确噪声点检测步骤的滤波算法,通过设定阈值并考虑相邻像素的相关性来区分噪声点和信号点,提高滤波精度。实验表明该算法在滤除噪声并保护图像细节方面比其他算法有较大提高,在严重噪声污染情况下,对图像的恢复也有较好效果。     

基于邻域结构相似性的混合噪音线性滤波算法

本文提出一种基于像素邻域结构信息相似性的混合噪音线性滤波算法(GLMF)。该算法是对线性混合滤波器(LMF)的一种改进,它利用图像中存在着大量冗余信息的特性,恢复被混合噪音染污的像素,在判断邻域内像素的相似性时,除考虑像素灰度值的相似性之外,又考虑了像素邻域结构的相似性,用像素灰度值的梯度来表示邻域结构信息。仿真实验证明,用GLMF去噪的视觉效果和峰值信噪比(PSNR)均优于已知的同类滤波器。该算法适用于恢复被高斯噪音和随机脉冲噪音混合污染的数字图像。     

图象椒盐噪声的非线性自适应滤除

为了在滤除图象椒盐噪声的同时能很好地保持图象的细节,提出了一种新颖的图象椒盐噪声非线性自适应滤除算法。该方法首先在噪声图象的滤波窗口中去除具有最大和最小灰度值的象素,然后求取剩余象素的均值,计算出该均值与对应的象素灰度值的差值,再通过与阈值相比较,确定是否用求得的均值代替原噪声图象的灰度值。阈值由图象的灰度分布自适应地确定,该算法与已发表的同类算法相比,具有更好地滤波性能,尤其在噪声严重时,其效果明显优于传统的中值滤波算法。