Pulse coupled neural network based anisotropic diffusion method for 1/f noise reduction

This paper describes that under certain operations, the parallel pulse transmission behavior of unit-linking PCNN (pulse coupled neural network) for image processing, which we have developed from PCNN based on biological experimental results, is equal to the open operation of mathematic morphology. The open operation of unit-linking PCNN based on its parallel pulse transmission behavior can be efficiently used for image processing, such as reducing the pulse noise of binary images. A new pulse-noise-reducing approach based on the open operation of unit-linking PCNN is proposed in this paper. The results of computer simulations show that the visual effects of retrieval images of binary pulse-noise images based on the open operation of unit-linking PCNN are better than those based on the median filter, a tradition approach for reducing image pulse noise. For binary images polluted by any other kinds of noise, such as white Guassian noise, after having turned the original noisy images into binary noisy images, using the approach proposed in this paper also can reduce the noise. This paper bridges the gap between image processing based on mathematical morphology and image processing based on unit-linking PCNN.

After summarizing the characteristics of pulse noise, a novel de-noising algorithm combined pulse coupled neural network (PCNN) with mathematical morphology was proposed in this paper. Firstly, noisy points were located using the synchronous pulse burst property of PCNN, then, were processed using the morphological open-close filtering. Finally our algorithm was compared with median filtering, morphological open-close filtering and the algorithm combined PCNN with median filtering. The experimental results showed that the proposed algorithm performed well, which was at a cost of much more calculating time than median filtering and morphological filtering, and cost far less than the algorithm combined PCNN with median filtering.

We propose a novel approach for the reduction of photon shot noise in image sensors, inspired by the M‐transform method. In the proposed method, the photon shot noise is transferred to small‐amplitude random signals. Then the small‐amplitude random signals are removed by an ϵ filter. The Sobel method was used for edge detection, and edges were preserved by not applying the ϵ filter on the edge. After noise reduction, the signals are restored to the image by a reverse M‐transform. We compared our method with the smoothing filter, the median filter, and the Wiener filter. In photon shot noise based on the Poisson distribution, we changed the number of incident photons and created the noise image by using a simulation. In cases where the maximum number of incident photons was 128 (illumination of about 2 lux), the proposed technique was subjectively confirmed as better than the smoothing filter, based on the effect of edge preservation even in a dark environment. As a result, the proposed method is confirmed to be effective for reducing photon shot noise. © 2011 Wiley Periodicals, Inc. Electron Comm Jpn, 95(1): 55–62, 2012; Published online in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/ecj.10344

Multi-frequency phase sensitive demodulation plays an important role in electrical impedance tomography (EIT). The phase sensitive detector is an effective tool for discriminating the amplitude and phase of each frequency component signal. It can be considered as the optimal linear filter in case of Gaussian-type noise. In real applications, however, the noise has also rather distinctive impulsive characteristics. It is a challenging task to remove the impulsive noise effectively in multi-frequency phase sensitive demodulation. In this paper, an approach for impulsive noise removal is presented in multi-frequency phase sensitive demodulation, which takes advantage of the periodicity of a multi-frequency signal and the significant properties of the median filter. The current work focuses on the use of two-dimensional (2D) median filter to handle the impulsive noise that existed in the measured signal, i.e. a 2D median filter is employed as the pre-stage signal conditioning module prior to the phase sensitive detector. The main advantages of the proposed method are that: 1) increasing the number of samples within the median filter window, which benefits the consecutive impulsive noise removal, and 2) no additional storage and computational resources are required for hardware system. Two classical methods are selected as comparisons in experiment. Numerical simulations and experimental results demonstrate that the proposed method can effectively eliminate the influence of impulsive noise on multi-frequency phase sensitive demodulation, and is able to achieve outstanding signal-to-noise ratio (SNR) performance.

Several adaptive order statistic filters (OSF) are developed and compared for channel characterization and noise suppression in images. Emphasis has been put on the situation when a noise-free reference image is not available but instead we can have a sequence of two noisy output images of the same input image through the communication channel twice. One of the noisy output images is used as the reference in the OSF. It is shown theoretically that if noises are not correlated, the expected values of the derived filter coefficients will be equal to those coefficients derived using a noise-free reference. Experiments using noisy reference image yield comparable results to those methods using a noise-free reference image and also better results than those of median, Gaussian, averaging and Wiener filters.

An adaptive filter was proposed to detect and remove pepper and salt noise in an image based on Pulse Coupled Neural Network(PCNN) firing matrix.The PCNN was simplified and a unidirectional decaying threshold was proposed to avoid complex parameter selection and improve processing speed.The noise-polluted pixels were detected through analyzing the PCNN firing matrix,and then only the noisy pixels were filtered by a median filter while protecting image edges and details.The window size of the filter and the filtering time were adaptively determined by calculating the noise intensity of the contaminated image.The experimental results show that the proposed method performs better in removing noise while conserving detailed information than traditional filters do.

Digital images are often corrupted by various types of noises amongst which impulse noise is most prevalent. Impulse noise appears during transmission and/or acquisition of images. Intrusion of impulse noise degrades the quality of the image and causes the loss of fine image details. Reducing the effect of impulse noise from corrupted images is therefore considered as an essential task to be performed before letting the image for further processing. However, the process of noise reduction from an image should also take proper care towards the preservation of edges and fine details of an image. A number of efficient noise reduction algorithms have already been proposed in the literature over the last few decades which have nurtured this issue with utmost importance. Design and development of new two dimensional (2D) filters has grown sufficient interest amongst the researchers. This chapter attempts to throw enough light on the advancement in this field by illustratively describing existing state-of-the-art filtering techniques along with their capability of denoising impulse noises.

Digital filtering is essential for digital imaging, image recognition, and super-resolution technology. For example, the presence of noise in images captured by digital cameras causes deterioration of the image quality and image recognition rate. In order to improve the image recognition rate, noise reduction and edge preservation must be performed during preprocessing. Noise is generally reduced using low-pass filters, such as the Gaussian filter. Although they reduce noise, such filters also have the properties of blurring edge. A strong edge blur reduces the accuracy of the feature detection in image recognition. Therefore, in our previous study, a fast M-estimation Gaussian filter for images (FMGFI) was proposed as an image filter that simultaneously achieves denoising and edge preservation. In the FMGFI, the setting of the optimal basic width of the 2nd order B-spline basis functions is important for achieving simultaneous denoising and edge preservation. In this method, the optimal basic width of the FMGFI was determined not only by manually setting the basic width but also by human judgment of the filtered images. Consequently, the inability to automatically determine the optimal basic width hindered efficient denoising during image processing Therefore, in this research, we develop and propose a method that can automatically determine the optimal basic width of the FMGFI. The previously proposed method calculates using the same basic width for all the pixels over the entire image; in contrast, the proposed method calculates using the basic width automatically determined for each pixel. The experiments confirmed that the method proposed in this study achieves higher denoising and edge preservation performance than the ones used in previous research. The results also showed that it has the highest denoising performance against salt-and-pepper noise as compared to other filters: non-local mean filter, Gaussian filter, median filter, bilateral filter, adaptive bilateral filter, and FMGFI. The experimental results for the Gaussian noise sowed that the proposed method has the same denoising and edge preservation performance as the other filters in visual evaluation. From the above, the proposed method is expected to contribute to efficient denoising and improvement of image quality by using it as a preprocessing.

Speckle noise generated during optical coherence tomography (OCT) imaging severely images image quality. In order to study the processing method of speckle noise reduction, this paper takes the OCT fundus image as the research object, and uses the median filtering, wavelet threshold method, Wiener filtering and anisotropic diffusion method to denoise the OCT fundus image. Finally, the effects of image noise reduction are objectively compared by four factors: signal-to-noise ratio (SNR), speckle index (Speckle-index), equivalent visual number (ENL) and edge retention coefficient (EPI). The results show that Wiener filtering and anisotropic diffusion method have better noise reduction effects.

Impulsive noise reduction is a special problem raised in active noise control (ANC) systems. The filtered-x least mean square (FxLMS) algorithm is a typical ANC method that has been successfully applied in practical applications. Some variants such as the logarithmic FxLMS and modified FxLMS algorithms have been proposed to deal with the impulsive noise for ANC. However, those methods require considerable computational loading in real-time applications. In this paper, a modified median FxLMS method is proposed for its low complexity to avoid the effect of the impulsive noise. Simulation results show that the proposed algorithm has better averaged noise reduction performance than the conventional FxLMS and other order statistics based FxLMS algorithms. In addition, the computational complexity stays the minimum requirement through using only the ordering operation, which is very suitable for the real-time implementation.

The problem of noise reduction has attracted a considerable amount of research attention over the past several decades. Among the numerous techniques that were developed, the optimal Wiener filter can be considered as one of the most fundamental noise reduction approaches, which has been delineated in different forms and adopted in various applications. Although it is not a secret that the Wiener filter may cause some detrimental effects to the speech signal (appreciable or even significant degradation in quality or intelligibility), few efforts have been reported to show the inherent relationship between noise reduction and speech distortion. By defining a speech-distortion index to measure the degree to which the speech signal is deformed and two noise-reduction factors to quantify the amount of noise being attenuated, this paper studies the quantitative performance behavior of the Wiener filter in the context of noise reduction. We show that in the single-channel case the a posteriori signal-to-noise ratio (SNR) (defined after the Wiener filter) is greater than or equal to the a priori SNR (defined before the Wiener filter), indicating that the Wiener filter is always able to achieve noise reduction. However, the amount of noise reduction is in general proportional to the amount of speech degradation. This may seem discouraging as we always expect an algorithm to have maximal noise reduction without much speech distortion. Fortunately, we show that speech distortion can be better managed in three different ways. If we have some a priori knowledge (such as the linear prediction coefficients) of the clean speech signal, this a priori knowledge can be exploited to achieve noise reduction while maintaining a low level of speech distortion. When no a priori knowledge is available, we can still achieve a better control of noise reduction and speech distortion by properly manipulating the Wiener filter, resulting in a suboptimal Wiener filter. In case that we have multiple microphone sensors, the multiple observations of the speech signal can be used to reduce noise with less or even no speech distortion.

In order to keep the detail information of edges while suppressing noise in laser active imaging, hybrid filter based on pixel classification is proposed. Based on contour curvature, the pixels are divided into different classes. Lee filter is used to suppress speckle noise and fuzzy weighted averaging filter is used to suppress other mixed noise. A contour point screening method is proposed to select candidate contour points from noisy image. Then the contours are thinned into single pixel. After eliminating the bogus contours brought by the noise and those that are comparatively week, the contour curvatures are calculated. According to the contour curvature, the pixels of the image are divided into different classes. For the pixels belonging to different classes, fuzzy weighted averaging filter and Lee filter with different filter parameters are used, which makes the proposed method own pixel-level adaptability. Experimental results indicate that the proposed method outperform the classical Lee, Kuan and wavelet soft-thresholding filter in denoising and keeping detail edge information.

Electrocardiograms (ECGs) can be conveniently obtained using capacitive ECG sensors. However, motion noise in measured ECGs can degrade R peak detection. To reduce noise, properties of reference signal and ECG measured by the sensors are analyzed and a new method of active noise cancellation (ANC) is proposed in this study. In the proposed algorithm, the original ECG signal at QRS interval is regarded as impulsive noise because the adaptive filter updates its weight as if impulsive noise is added. As the proposed algorithm does not affect impulsive noise, the original signal is not reduced during ANC. Therefore, the proposed algorithm can conserve the power of the original signal within the QRS interval and reduce only the power of noise at other intervals. The proposed algorithm was verified through comparisons with recent research using data from both indoor and outdoor experiments. The proposed algorithm will benefit a noise reduction of noisy biomedical signal measured from sensors.

Noise Suppression from images is one of the most important concerns in digital image processing. Impulsive noise may occur during image acquisition, transmission or storage Noise should be removed in such a way that important information of image should be preserved. We can use so many algorithms for getting the original image, by removing salt and pepper noise from the corrupted images. In this paper an algorithm is proposed for the restoration of gray scale images that are highly corrupted by impulse noise (salt and pepper noise). There are two phases in the proposed algorithm. First phase detects whether the processing pixel is corrupted or not. In the Second phase it recreates the corrupted pixel by means of the proposed algorithm. This algorithm shows better results than the Standard Median Filter (MF), Center Weighed Median Filter(CWM), Adaptive Median Filter(AMF), Adaptive Center Weighed Median Filter (ACWM), Decision Based Algorithm (DBA) and Modified Decision Based Unsymmetrical Trimmed Median Filter Algorithm(MDBUTMF). Obtained results with different grayscale images shows that proposed algorithm gives better Peak Signal-to-Noise Ratio (PSNR) and less Computational time and works well in removing salt and pepper noise at low, medium and high noise densities.

Restoration of original image corrupted with high density salt-and-pepper noise is still a challenging task. In this letter, we propose here a new method; Multi-Pass Unsymmetric Trimmed Median Filter (MPUTMF) [versions ‘a’ and ‘b’], to restore an image affected with high density salt-and-pepper noise, with better edge preservation. The MPUTMFa can do the restoration within two passes over the noisy image using the preprocessed pixels obtained in the same pass, where as MPUTMFb can take up to six passes over the noisy image for restoration without using the preprocessed pixels. MPUTMFb is computationally efficient on single core processor systems where as MPUTMFb is well suited to be implemented on parallel processing systems or GPUs to achieve higher computational efficiency. The proposed methods are compared with conventional as well as advanced algorithms like Median Filter (MF), Adaptive Median Filter (AMF), Efficient Decision Based Algorithm (EDBA), Improved Decision Based Algorithm (IDBA) and Modified Decision Based Unsymmetric Trimmed Median Filter (MDBUTMF). The experimental analysis (visual and quantitative) shows that our method gives better results on images affected with high density salt-and-pepper noise. Peak Signal-to-Noise Ratio (PSNR) and Image Enhancement Factor (IEF) are used for quantitatively evaluating the results of proposed algorithm(s).KeywordsImage restorationSalt-and-pepper noiseMedian filterTrimmed median filterImpulse noiseParallel processing