Edge-Aware Level Set Diffusion and Bilateral Filtering Reconstruction for Image Magnification

In this paper we propose an image magnification reconstruction method.In recent years many interpolation algorithms have been proposed for image magnification,but all of them have defects to some degree,such as jaggies and blurring.To solve these problems,we propose applying post-processing which consists of edge-aware level set diffusion and bilateral filtering.After the initial interpolation,the contours of the image are identified.Next,edge-aware level set diffusion is applied to these significant con...     

一种图像放大的偏微分方程综合模型

针对传统图像放大处理过程中基于线性插值方法通常导致边缘模糊问题,分析了各向同性扩散模型和各向异性扩散模型在图像处理中的优缺点,提出了一种线性扩散和P-M方程自适应结合的图像放大综合模型。该模型对图像非平滑区域采用各向异性扩散模型处理,而平滑区域则采用各向同性扩散模型处理。实验结果表明,该综合模型在保持图像边缘锐度的同时提高了图像的清晰度,能够有效提高放大图像的主观视觉质量和客观SNR及PSNR。     

热扩散和整体变分模型自适应调整的图像放大模型

近年来基于偏微分方程的图像处理因其具有更强的局部自适应特性和高度的灵活性而成为继小波图像处理之后新的研究热点。本文在对热扩散模型和TV（Total Variation）模型分析的基础上,针对二模型在图像平滑区域及边缘或纹理区域扩散所表现出的不同特性和优势,提出一种基于图像局部特性的自适应混合图像放大模型,该模型以双线性插值后的放大图像为初始值,根据像素点梯度特性在平滑区域和边缘或纹理区域自适应的调整扩散模型,使得在平滑区域进行热传导的同性扩散,而在边缘和纹理区域处则进行沿与梯度方向相垂直的方向扩散。该模型很好的抑制了图像插值放大后所带来的块状效应,以及放大后边缘或纹理区域的假边缘现象。理论上证明了所提出模型解的存在性,同时大量的仿真实验验证了所提出模型的有效性。     

Quaternion Diffusion for Color Image Filtering

How to combine color and multiscale information is a fundamental question for computer vision, and quite a few color diffusion techniques have been presented. Most of these proposed techniques do not consider the direct interactions between color channel pairs. In this paper, a new method of color diffusion considering these effects is presented, which is based on quaternion diffusion （QD） equation. In addition to showing the solution to linear QD and its analysis, one form of nonlinear QD is discussed. Compared with other color diffusion techniques, considering the interactions between channel pairs, QD has the following advantages： 1） staircasing effect is avoided; 2） as diffusion tensor, the image derivative is regularized without requiring additional convolution; 3） less time is needed. Experimental results demonstrate the effectiveness of linear and nonlinear QD applied to natural color images for denoising by both visual and quantitative evaluations.     

自适应耦合方向扩散和冲击滤波的图像放大

提出一种新的基于偏微分方程（PDE）的图像放大算法。算法模型由方向扩散、保真项和冲击滤波构成。通过设置与边缘信息有关的保真项系数,调节图像被模糊的程度,保留图像特征;综合调节扩散方程和冲击滤波的权重,实现两种PDE的自适应耦合。由于运用连续函数,结合图像内容调节两种PDE的权重,避免了在放大时出现人工痕迹。实验结果表明,此算法在客观指标和主观效果上均优于传统双线性插值和其他同类方法,是一种有效的图像放大算法。     

Reversible Interpolation of Vectorial Images by an Anisotropic Diffusion-Projection PDE

In this paper, a nonlinear model for the interpolation of vector-valued images is proposed. This model is based on an anisotropic diffusion PDE and performs an interpolation that is reversible. The interpolation solution is restricted to the subspace of functions that can recover the discrete input image, after an appropriate smoothing and sampling. The proposed nonlinear diffusion flow lies on this subspace while its strength and anisotropy adapt to the local variations and geometry of image structures. The derived method effectively reconstructs the real image structures and yields a satisfactory interpolation result. Compared to classic and other existing PDE-based interpolation methods, our proposed method seems to increase the accuracy of the result and to reduce the undesirable artifacts, such as blurring, ringing, block effects and edge distortion. We present extensive experimental results that demonstrate the potential of the method as applied to graylevel and color images. The authors acknowledge the financial support of the Future and Emerging Technologies (FET) programme ‘ASPI’ within the Sixth Framework Programme for Research of the European Commission, under FET-Open contract No. 021324.     

一种基于非线性扩散方程的PET重建方法

正电子发射断层重建是一类不适定问题,因此正则化方法常用来抑制噪声提高重建的质量.依据Mumford-Shah泛函模型提出了该类重建问题的统一变分模型.基于这一模型,提出了一种新的基于扩散方程的PET重建算法,同时给出了线性和非线性两种方法.实验结果表明,两种方法均有良好的重建性能,并且非线性方法通过合理的设置参数,能体现出较为优越的抗噪声性能和边缘保持特性.     

结合中值滤波的彩色自蛇模型在遥感图像放大的研究

研究了彩色自蛇模型的形成过程,对遥感TM图像进行了后处理放大处理,采用最邻近点插值、双线性插值、三次样条插值、双立方插值、双二次插值。针对插值放大后的图像的边缘出现锯齿化和模糊化问题,由于彩色自蛇模型本身不仅具有边缘锐化功能,它可以消除边缘锯齿化和边缘模糊化问题,而且具有去噪的能力。采用彩色自蛇模型进行后处理,并对处理后的残余斑点进行研究,结果采用了中值滤波的方式取得更好的效果,并对含噪的遥感图像也有较好的放大效果,实验结果证明该方法适于遥感图像的放大处理。     

梯度角约束图像插值

为了使插值的图像既保持光滑的图像轮廓又获得清晰的图像边缘,提出一种梯度角约束图像插值方法.首先证明了局部恒定的梯度角约束插值等价于总变分最小化;在此基础上,提出一个具有更强连续性的梯度角约束,由这个约束条件导出的三阶偏微分方程使插值的图像既可以保证水平线方向的连续性,又减小图像边缘的宽度;最后证明了插值方程中正则项的对比不变性.通过数值实验证明了该方法的有效性.     

基于GPU的血管造影图像增强方法*

血管增强扩散算法遵循多尺度方法,利用非线性各向异性扩散方法进行血管增强,该方法在可视化不同半径的血管和增强血管外观上比现存的大部分方法都要好,但医学图像数据分辨率和灰度级都很高,多尺度选择和求解非线性各向异性扩散的偏微分方程时运算量很大,执行速率低,不适合实际应用。提出一种基于GPU(graphic processing unit)的血管造影图像增强方法,采用计算统一设备架构(CUDA)技术,利用像素的独立性和偏微分方程求解的并发性,实现了并行血管增强扩散算法。实验结果表明,该方法在保持血管增强效果一样的同时降低了处理时间,加速比达到27倍以上。     

Improved Geometric Anisotropic Diffusion Filter for Radiography Image Enhancement

In radiography imaging, contrast, sharpness and noise there are three fundamental factors that determine the image quality. Removing noise while preserving and sharpening image contours is a complicated task particularly for images with low contrast like radiography. This paper proposes a new anisotropic diffusion method for radiography image enhancement. The proposed method is based on the integration of geometric parameters derived from the local pixel intensity distribution in a nonlinear diffusion formulation that can concurrently perform the smoothing and the sharpening operations. The main novelty of the proposed anisotropic diffusion model is the ability to combine in one process noise reduction, edge preserving and sharpening. Experimental results using both synthetic and real welding radiography images prove the efficiency of the proposed method in comparison with other anisotropic diffusion methods.     

TV图像插值的双方向扩散改进算法

提出了一种双方向扩散图像插值方法,有效地减小了插值图像的边缘宽度,从而获得清晰而光滑的图像边缘.这种双方向扩散在图像边缘斜坡较亮一侧进行前向扩散,而在边缘斜坡较暗一侧进行后向扩散.同时,它能根据图像边缘的特征自适应地调整前向、后向扩散强度,从而避免了在插值的图像中产生虚的纹理或边缘.数值实验结果显示,该方法既能很好地减小插值图像的边缘宽度,又不会产生其他人工虚像.     

三次样条插值的逆扩散图像增强算法

提出了非线性逆扩散图像增强算法的原则，应用三次样条插值方法构造了该算法的扩散系数，得到了较好的图像增强效果。说明该算法能够对有微弱噪声的图像进行图像增强并保持细节纹理，其扩散系数既满足了逆扩散图像增强的原则，又容易构造。     

基于小波变换的图象放大方法再探讨

图象放大 (又称图象变焦、图象重采样 )问题严格地说是一个病态问题 ,根据不同的模型 ,人们先后提出了多种图象放大方法 ,如线性插值、三次样条插值、分形插值以及基于小波变换的方法等 .图象放大问题的焦点是如何在图象放大过程中保持良好的视觉分辨率 ,表现在基于小波变换的图象放大方法中就是如何构造图象高频分量的问题 .针对目前常用的变换域内插方法 ,在一维信号上所作的模拟实验表明 ,该方法并不合理 .对常用的几种图象放大效果评价的方法进行了比较分析 ,结果认为 ,最理想方法还是多人主观评判法 .由于小波高频系数构建问题并未有理想方案 ,因此 ,对小波基函数的选择问题必须作进一步研究 .     

结合异质扩散和小波包分解的医学图像分割算法

提出一种医学图像解剖结构特征提取算法,结合异质扩散和小波包分解实现了医学图像分割.在异质扩散基础上,提出一种新的FAB扩散方法对图像进行去噪,同时保持图像解剖结构的边缘和局部细节;然后用一个基于小波包分解的自适应边缘检测算法来检测图像的解剖结构的轮廓特征.分割结果表明,该方法能在有效地去除噪声的基础上完成医学图像解剖结构轮廓特征的提取.     

Fast Anisotropic Smoothing of Multi-Valued Images using Curvature-Preserving PDE's

We are interested in PDE's (Partial Differential Equations) in order to smooth multi-valued images in an anisotropic manner. Starting from a review of existing anisotropic regularization schemes based on diffusion PDE's, we point out the pros and cons of the different equations proposed in the literature. Then, we introduce a new tensor-driven PDE, regularizing images while taking the curvatures of specific integral curves into account. We show that this constraint is particularly well suited for the preservation of thin structures in an image restoration process. A direct link is made between our proposed equation and a continuous formulation of the LIC's (Line Integral Convolutions by Cabral and Leedom (1993). It leads to the design of a very fast and stable algorithm that implements our regularization method, by successive integrations of pixel values along curved integral lines. Besides, the scheme numerically performs with a sub-pixel accuracy and preserves then thin image structures better than classical finite-differences discretizations. Finally, we illustrate the efficiency of our generic curvature-preserving approach – in terms of speed and visual quality – with different comparisons and various applications requiring image smoothing : color images denoising, inpainting and image resizing by nonlinear interpolation.     

基于多进制小波变换的图象放大方法

多进制小波是近几年发展的小波分析理论的一个新分支，为了更好地进行图象的放大，提出了一个基于多进制小波变换的图象放大方法，并用三进制小波变换进行了图象的放大试验，通过与常用的插值放大方法进行比较的结果表明，基于多进制小波放大的图象能较好地保持原来图象的特征，优于一般的插值放大方法。     

图像的定向插值放大及其简化算法

首先介绍了常见的图像插值放大方法及其不足 ,然后描述了定向插值放大的基本思想和实现方法 ,最后分析了所提到的插值放大方法的优缺点及各自的适应范围 ,并讨论了如何进一步地提高插值放大算法的性能     

保持轮廓清晰光滑的灰度图像放大算法

灰度图像放大时，插值所具有的平滑作用会退化图像的高频部分，使放大图像轮廓变得模糊，文中提出一种基于拟合分界线的插值放大算法，该处包括分割和插值放大两个步骤；分割是搜索出灰度图像的突变象素点，并用三次均匀B样条把它们拟合为光滑分界线，以把整幅图像分割为若干子区域；插值放大是基于拟合分界线对图像插值，即插值操作限定于原图像的某一子区域内进行。采用文中算法得到的放大图像不仅可保持轮廓清晰，而且可持续轮廓光滑，最后给出三个放大实例，证明了 文中放大算法比常规算法产生的图像质量高。     

快速方向字典插值的中低倍率超分辨率算法

在中低倍率实时超分辨率显示系统中,为了进一步提高重建图像质量的同时降低运算复杂度,提出了一种全方向字典拟合插值的中低倍率快速超分辨率算法。对每个高分辨率图像中待插值像素,首先在对应的低分辨率插值区间中,采用金字塔纹理字典查找主纹理方向,然后根据待插值像素位置进行插值方向修正,最后通过单边拟合的插值方法得到插值结果。实验结果表明,与基于线性映射的边缘定位超分辨率算法(SREO)相比,该算法的峰值信噪比(PSNR)和结构相似度(SSIM)平均分别提高了0.73 dB和0.04,同时平均运算时间减少了40%。