Noise reduction

About: Noise reduction is a research topic. Over the lifetime, 25121 publications have been published within this topic receiving 300815 citations. The topic is also known as: denoising & noise removal.

A Fuzzy Noise Reduction Method for Color Images

Abstract: A new fuzzy filter is presented for the reduction of additive noise for digital color images. The filter consists of two subfilters. The first subfilter computes fuzzy distances between the color components of the central pixel and its neighborhood. These distances determine in what degree each component should be corrected. All performed corrections preserve the color component distances. The goal of the second subfilter is to correct the pixels where the color components differences are corrupted so much that they appear as outliers in comparison to their environment. Experimental results show the feasibility of the proposed approach. We compare with other noise reduction methods by numerical measures and visual observations. We also illustrate the performance of the proposed method as preprocessing step for edge detection

A two-step noise reduction technique

Abstract: The paper addresses the problem of single microphone speech enhancement in noisy environments Common short-time noise reduction techniques proposed in the art are expressed as a spectral gain depending on the a priori SNR In the well-known decision-directed approach, the a priori SNR depends on the speech spectrum estimation in the previous frame As a consequence, the gain function matches the previous frame rather than the current one which degrades the noise reduction performance We propose a new method, called the two-step noise reduction (TSNR) technique, which solves this problem while maintaining the benefits of the decision-directed approach This method is analyzed and results in voice communication and speech recognition contexts are given

Clinical evaluation of signal-to-noise ratio-based noise reduction in Nucleus® cochlear implant recipients.

Abstract: Objective: The aim of this study was to investigate whether a real-time noise reduction algorithm provided speech perception benefit for Cochlear™ Nucleus® cochlear implant recipients in the laboratory. Design: The noise reduction algorithm attenuated masker-dominated channels. It estimated the signal-to-noise ratio of each channel on a short-term basis from a single microphone input, using a recursive minimum statistics method. In this clinical evaluation, the algorithm was implemented in two programs (noise reduction programs 1 [NR1] and 2 [NR2]), which differed in their level of noise reduction. These programs used advanced combination encoder (ACE™) channel selection and were compared with ACE without noise reduction in 13 experienced cochlear implant subjects. An adaptive speech reception threshold (SRT) test provided the signal-to-noise ratio for 50% sentence intelligibility in three different types of noises: speech-weighted, cocktail party, and street-side city noise. Results: In all three noise types, mean SRTs for both NR programs were significantly better than those for ACE. The greatest improvement occurred for speech-weighted noise; the SRT benefit over ACE was 1.77 dB for NR1 and 2.14 dB for NR2. There were no significant differences in speech perception scores between the two NR programs. Subjects reported no degradation in sound quality with the experimental programs. Conclusions: The noise reduction algorithm was successful in improving sentence perception in speech-weighted noise, as well as in more dynamic types of background noise. The algorithm is currently being trialed in a behind-the-ear processor for take-home use.

Using local median as the location of the prior distribution in iterative emission tomography image reconstruction

Abstract: Iterative reconstruction algorithms like MLEM (Maximum Likelihood Expectation Maximization) can be regularized using a weighted roughness penalty term according to certain a priori assumptions of the desired image. In the R?RP (Median Root Prior) algorithm the penalty is set according to the deviance of a pixel from the local median. This allows both noise reduction and edge preservation. The prior distribution is Gaussian located around the median of a neighborhood of the pixel. Non-monotonic details smaller than a given limit are considered as noise and are penalized. Thus, MRP implicitly contains the general description of the characteristics of the desired emission image, and good localization of tissue boundaries is achieved without anatomical data. In contrast to the MLEM method, the number of iterations needs not be restricted and unlike many other Bayesian methods MRP has only one parameter. The penalty term can be applied to various iterative reconstruction algorithms. The assumption that the true pixel value is close to the local median applies to any emission images, including the 3D acquisition and images reconstructed from parametric sinograms.