Yi Zhan

Nonlocal means method using weight refining for despeckling of ultrasound images

Speckle noise is an inherent nature of ultrasound images, which have negative effect on image interpretation and diagnostic tasks. In this paper, a nonlocal means method using weight refining for ultrasonic speckle reduction is proposed. Based on a signal-dependent speckle model, a novel similarity weight is derived by Bayesian framework. The weight is iteratively refined in a lower dimensional subspace using principal components analysis (PCA) to improve accuracy of weight and reduce its computational complexity. The weight refining is automatically terminated using mean absolute error based on a fully formed speckle region estimated by a PCA-based method. Simulations on various images demonstrate that our method can provide significant improvement over other evaluated methods. Thus, our method has great potential applications to medical ultrasound imaging. HighlightsAn extension of nonlocal means method to ultrasonic speckle reduction.Weight refining scheme in a lower dimensional PCA subspace.Automatic termination of weight refining scheme using mean absolute error based on an estimated fully formed speckle region.Superior restoration performance compared with existing ultrasound image despeckling methods.Great potential applications to medical ultrasound imaging.

Fast communication: Decision-based non-local means filter for removing impulse noise from digital images

The decision-based non-local means filter is proposed to remove fixed-value impulse noise from the corrupted digital images. The proposed filter first identifies the corrupted pixels using the local statistics based noise detector and then removes the detected impulses using the reference image-based non-local means filter while keeping the uncorrupted pixels unaltered. Extensive simulations demonstrate that the proposed filter can remove impulse noise from the corrupted images effectively while preserving image details very well at the various noise ratios, which leads to its significantly better image restoration performance than numerous state-of-the-art switching-based filters.

A novel iterative non-local means algorithm for speckle reduction

Despeckling of ultrasound images is a crucial step for facilitating subsequent image processing. The non-local means (NLM) filter has been widely applied for denoising images corrupted by Gaussian noise. However, the direct application of this filter in ultrasound images cannot provide satisfactory restoration results. To address this problem, a novel iterative adaptive non-local means (IANLM) filter is proposed to despeckle ultrasound images. In the proposed filter, the speckle noise is firstly transformed into additive Gaussian noise by square root operation. Then the decay parameter is estimated based on a selected homogeneous region. Finally, an iterative strategy combined with the local clustering method based on pixel intensities is adopted to realize effective image smoothing while preserving image edges. Comparisons of the restoration performance of IANLM filter with other state-of-the-art despeckling methods are made. The quantitative comparisons of despeckling synthetic images based on Peak signal-to-noise ratio (PSNR) show that the IANLM filter can provide the best restoration performance among all the evaluated filters. The subjective visual comparisons of the denoised synthetic and ultrasound images demonstrate that the IANLM filter outperforms other compared algorithms in that it can achieve better performance of noise reduction, artifact avoidance, edges and textures preservation and contrast enhancement.

Nonlocal-means-based smallest univalue segment assimilating nucleus edge detector

Abstract. To improve the antinoise performance of the smallest univalue segment assimilating nucleus (SUSAN) edge detector, a nonlocal means-based SUSAN edge detector is proposed. The proposed method first determines the initial SUSAN edge response based on the image patch convolved with an adaptive kernel instead of the single pixel. Then it computes the final edge response using the weighted sum of the initial edge responses of the pixels with their structures similar to the considered pixel. Extensive simulations on natural and real images demonstrate that compared with state-of-the-art detectors, the proposed method performs much better in terms of robustness to noise and edge detection and it provides significantly higher values of Pratt’s figure of merit and performance measure.

Pixel-wise decay parameter adaption for nonlocal means image denoising

Abstract. The globally fixed decay parameter is generally adopted in the traditional nonlocal means method for similarity computation, which has a negative influence on its restoration performance. To address this problem, we propose to adaptively tune the decay parameter for each image pixel using the golden section search method based on the pixel-wise minimum mean square error, which can be estimated using the prefiltered result and the estimated noise component. The quantitative and subjective comparisons of restoration performance among the proposed method and several state-of-the-art methods indicate that it can achieve a better performance in noise reduction, artifact avoidance, and detail preservation.

An Improved Non-local Means Filter for Image Denoising

This paper proposes a improved non-local means (NLM) filter for image denoising. Due to the drawback that the similarity is computed based on the noisy image, the traditional NLM method easily generates the artifacts in case of high-level noise. The proposed method first preprocesses the noisy image by Gaussian filter. Then, a moving window at each pixel of the noisy image is chosen as the search window, and meanwhile, a improved calculation method of spatial distance based on the preprocessed image is used for computing the similarity. Finally, combining the improved distance with search window based on the noisy image, the intensity of each pixel is restored as the traditional NLM method. The standard images are used to evaluate restoration performance of the proposed method. Additionally, the application on medical image denoising also demonstrates that our method is practical.

Spiking cortical model-based noise detector for switching-based filters

A novel noise detector based on the spiking cortical model (SCM) is proposed for switching-based filters. In the proposed noise detector, the corrupted pixels are firstly identified as noise candidates based on the firing time of the SCM, and then the misclassified noise-free pixels are dismissed from noise candidates based on the absolute difference of the firing time between the considered neurons and their neighboring neurons. Extensive simulations show that although the proposed noise detector generally has lower computational efficiency than several state-of-the-art noise detectors, it outperforms all the compared noise detectors in noise detection accuracy by classifying the pixels in the corrupted images with very few or no mistakes at the various noise ratios.