Shangbo Zhou

Fractional nonlinear anisotropic diffusion with p-Laplace variation method for image restoration

In this paper, a novel class of fractional-order nonlinear anisotropic diffusion equations based image restoration model is established, which employs the p-Laplace norm of fractional-order gradient of an image intensity function. The role of the fractional-order gradient is to better accommodate the texture details of an image, and the adaptive factor p can be used to diffuse adaptively according to local geometry features, which are fractional-order curvature and fractional-order gradient of an image. Besides removing noise and non-linearly keeping high-frequency edge of images efficiently, our proposed model can enhance the texture details of images and greatly eliminate the staircase effects and also the speckle effects. Fourier transform technique is also proposed to compute the fractional order derivative. Experimental results illustrate that our proposed model can deal with edge preserving and texture enhancing, more efficiently than the other four methods and outperform the other four methods by means of PSNR. Our average PSNR is closed to 1dB higher than the average PSNRs of the other four methods.

Fuzzy color distribution chart -based shot boundary detection

Shot boundary detection is an important research topic in the field of video processing technology, which has a wide range of applications in video indexing, pattern recognition, video summarization, video classification, video retrieval, etc. Shot boundary detection includes both abrupt (cut) and gradual transition detection. In this paper, a new method is proposed for extracting the feature from frames of a video. We name the proposed method as fuzzy color distribution chart (FCDC). FCDC can be used to describe the spatial distribution of colors and avoid the influences of noise, slight illumination and insertions such as words and logos. Based on the FCDC, a new algorithm is put forward for shot boundary detection, which can distinguish the gradual transition if there are quickly moving objects in the frames. Our proposed algorithm can be employed to suppress some defects of shot boundary detection that cannot be solved completely, and the experimental results show that the improved algorithm can detect the shot boundary more accurately than some existing researches.

Image Denoising Algorithm Based on Entropy and Adaptive Fractional Order Calculus Operator

In this paper, a fractional calculus operator for image denoising is constructed based on the characteristic of local entropy and the gradient feature, and an adaptive fractional calculus image denoising algorithm is proposed. First, the effects on the entropy and gradient by noise are analyzed, respectively. Second, the noise points are regarded as small probability events in an image, and the noise points, edges, texture regions, and smooth regions are divided combining with the local structure. Finally, for improving the image denoising effect, we consider employing different fractional orders to deal with different pixels and a piecewise function is constructed to make the differential order to be adaptive. The function is with respect to the local entropy and gradient on the pixel. The experimental results show that the peak signal-to-noise ratio and the entropy (ENTROPY) of the proposed adaptive fractional calculus image denoising algorithm are higher than that of the other algorithms compared in this paper. The proposed algorithm can not only preserve image edges and texture information while removing the noise, but also obtain a good visual effect.

A novel method for automatically locating the pylorus in the wireless capsule endoscopy.

Abstract Wireless capsule endoscopy (WCE) is a non-invasive technique used to examine the interiors of digestive tracts. Generally, the digestive tract can be divided into four segments: the entrance; stomach; small intestine; and large intestine. The stomach and the small intestine have a higher risk of infections than the other segments. In order to locate the diseased organ, an appropriate classification of the WCE images is necessary. In this article, a novel method is proposed for automatically locating the pylorus in WCE. The location of the pylorus is determined on two levels: rough-level and refined-level. In the rough-level, a short-term color change at the boundary between stomach and intestine can help us to find approximately 70–150 positions. In the refined-level, an improved Weber local descriptor (WLD) feature extraction method is designed for gray-scale images. Compared to the original WLD calculation method, the method for calculating the differential excitation is improved to give a higher level of robustness. A K-nearest neighbor (KNN) classifier is incorporated to segment these images around the approximate position into different regions. The proposed algorithm locates three most probable positions of the pylorus that were marked by the clinician. The experimental results indicate that the proposed method is effective.

Super-resolution image reconstruction method using homotopy regularization

In this paper, we present a homotopy regularization based on fractional order total variation for image super-resolution. This regularization function of the proposed method is composed of three parts: fractional order TV regularization term, homotopy data fidelity term and traditional data fidelity. And the overall function is minimized using the gradient descent flow. For super-resolution reconstruction, the proposed approach makes full use of characteristics of fractional calculus and homotopy method, avoiding effectively the jaggies when reconstructing an image. Fractional calculus can significantly improve high frequency component of a signal, while the low frequency part is strengthened accordingly. Homotopy method can achieve convergence quickly in a nonlocal area.

Range Limited Peak-Separate Fuzzy Histogram Equalization for image contrast enhancement

Histogram equalization is a well-known technique for enhancing image contrast for its simplicity and effectiveness. However, the existing approaches to this technique may change the contrast so sharply that it is unsuitable to be implemented in consumer electronics. In this paper, we propose a novel histogram equalization method referred to as Range Limited Peak-Separate Fuzzy Histogram Equalization (RLPSFHE), which aims to gain a good trade-off between mean-brightness preservation and contrast enhancement, so that it can be applied in consumer electronics. In the RLPSFHE, fuzzy statistics is applied to deal with digital images for their representation, and a set of peaks is calculated from the crisp fuzzy histogram, which is a set of points for separation. Since then, the input fuzzy histogram can be divided into several segments with those points of peak. After that, an intensity factor is employed to control the intension of brightness preservation when a range limited method is used to process each sub-histogram, the experimental results show that the RLPSFHE can achieve a better trade-off between mean-preservation and contrast enhancement.

Image Denoising Based on Adaptive Fractional Order with Improved PM Model

In order to improve the image quality, in this paper, we propose an improved PM model. In the proposed model, we introduce two novel diffusion coefficients and a residual error term and replace the integer differential operator with the fractional differential operator in the PM model. The diffusion coefficients can be used effectively for edge detection and noise removal. The residual error term can help to prevent image distortion. Fractional order differential operator has a good characteristic that it can enhance image texture information while removing image noise. Additionally, in the two new diffusion coefficients, a novel method is proposed for automatically setting parameter k, and it does not need to do any experiments to get the value of . For the computing fractional order diffusion coefficient, we employ the discrete Fourier transform, and an iterative scheme is carried out in the frequency domain. In the proposed model, not only is the integer differential operator replaced with the fractional differential operator, but also the order of the fractional differentiation is determined adaptively with the local variance. Comparing with some existing models, the experimental results show that the proposed algorithm can not only better suppress noise, but also better preserve edge and texture information. Moreover, the running time is greatly reduced.