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。     

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。     

Image magnification method using joint diffusion

In this paper a new algorithm for image magnification is presented. Because linear magnification/interpolation techniques diminish the contrast and produce sawtooth effects, in recent years, many nonlinear interpolation methods, especially nonlinear diffusion based approaches, have been proposed to solve these problems. Two recently proposed techniques for interpolation by diffusion, forward and backward diffusion (FAB) and level-set reconstruction (LSR), cannot enhance the contrast and smooth edges simultaneously. In this article, a novel Partial Differential Equations (PDE) based approach is presented. The contributions of the paper include: firstly, a unified form of diffusion joining FAB and LSR is constructed to have all of their virtues; secondly, to eliminate artifacts of the joint diffusion, soft constraint takes the place of hard constraint presented by LSR; thirdly, the determination of joint coefficients, criterion for stopping time and color image processing are also discussed. The results d     

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。     

一种新的基于小波变换的图像放大算法

图像放大技术的主要难点在于如何使放大后的图像尽可能保持原始图像的清晰度.传统的内插法存在着一定的缺陷.提出了一种新的图像放大算法,该算法以小波变换为基础,采用分段样条插值法进行高频系数放大,通过对变换系数进行非线性增强保证放大图像的清晰度.实验证明,该方法具有很好的放大效果.     

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。     

一类基于加权系数的插值算法

插值运算作为一种标准的运算在图象处理领域，尤其是在图象的放大、旋转和卷积处理过程中有着广泛的应用。当图象的轮廓被提取出来后，插值核的选取要考虑到计算速度和数学精度的要求。在本文中，提出了一种新的插值算法．引入了八个加权系数。通过它们可以得到一个加权公式。通过软件编程实现该算法后，处理后的图象效果得到了较好的提高，并且有效地减轻了锯齿现象。     

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 .     

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。     

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     

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。     

Image enlargement based on a step edge model

In compiling a multimedia document we often need to enlarge the size of an image. The traditional pixel repetition method tends to make the edges jagged. On the other hand, the interpolation-based methods tend to make the edges blurry in the enlarging process. In this paper we propose an image magnification method based on a step edge model. Using the model, we are able to define a straight step edge segment with four parameters. In enlarging a digital image, we first derive the parameter values for its step edge segments. This is like extracting the step edges in the corresponding continuous image. Then we re-digitize the step edges in the continuous image with a finer grid to obtain an enlarged image. In this way, the step edges are able to stay well defined after they are enlarged. The experiments show that in both visual comparison and quantitative analysis, the results produced by the suggested step edge model-based approach are consistently and significantly better than that produced by pixel repetition and bilinear interpolation.     

局部几何结构驱动的图像插值放大及超分辨率复原

众所周知,图像插值是根据一幅低分辨率噪声图像重建相应高分辨率清晰图像的数字图像处理技术。虽然已有一些文献报道了多种图像插值算法,然而现有算法在插值视觉效果和计算复杂度两者间往往难以实现均衡,为此,提出了一种局部几何结构驱动的偏微分方程(PDE)图像插值算法。该算法通过耦合边缘、纹理和角形3种不同几何结构的扩散机制来进行插值,插值结果表明,该算法不仅具有抗噪声性能,而且能够同时增强边缘、纹理以及角形结构。考虑到图像的超分辨率复原与插值放大在数学本质上的一致性,特将上述PDE应用推广到图像的超分辨率复原,并且针对高强度噪声情形下,超分辨率图像中出现的伪纹理结构,提出了一种耦合全变差模型的改进的PDE。实验结果表明,不论是插值放大图像,还是超分辨率复原图像都具有较高的视觉质量和峰值信噪比。     

Photographic Appearance Enhancement via Detail-Based Dictionary Learning

A number of edge-aware filters can efficiently boost the appearance of an image by detail decomposition and enhancement. However, they often fail to produce photographic enhanced appearance due to some visible artifacts, especially noise, halos and unnatural contrast. The essential reason is that the guidance and the constraint of high-quality appearance are not sufficient enough in the process of enhancement. Thus our idea is to train a detail dictionary from a lot of high-quality patches in order to constrain and control the entire appearance enhancement. In this paper, we propose a novel learningbased enhancement method for photographic appearance, which includes two main stages: dictionary training and sparse reconstruction. In the training stage, we construct a training set of detail patches extracted from some high-quality photos, and then train an overcomplete detail dictionary by iteratively minimizing an ? ₁-norm energy function. In the reconstruction stage, we employ the trained dictionary to reconstruct the boosted detail layer, and further formalize a gradient-guided optimization function to improve the local coherence between patches. Moreover, we propose two evaluation metrics to measure the performance of appearance enhancement. The final experimental results have demonstrated the effectiveness of our learning-based enhancement method.     

An Algorithm For Segmentation Under Interpolation Conditions Using Deformable Models

We present a deformable model technique for geophysical image analysis. Deformable model approaches have been developed extensively in the literature, including prior applications to geophysical or medical image interpretation. In this paper we propose a method to segment a geophysical image under interpolation conditions (well data). The originality of this segmentation method is that it considers the deformable model as a set of articulated curves, which corresponds to the interfaces between different regions. Moreover, the interpolation conditions permit some geometric constraints to be made on the model. The theoretical aspect of the method is given in the case of a three dimensional image. Numerical results are given.     

基于加权抛物线插值与小波变换的图像放大算法

图像放大问题的焦点是如何在图像放大过程中保持良好的视觉分辨率和清晰度。论文首先对加权抛物线插值算法进行了改进,在此基础上提出一种加权抛物线插值结合小波变换的图像放大算法,并对小波变换后的小波系数低频带作了幅值增强处理。实验结果表明,相对于传统的图像放大算法,该算法不仅较好地保留了原始图像的细节信息,而且还提高了放大后图像的亮度和清晰度。     

基于色度-亮度全变分模型的彩色图像放大

采用传统插值法放大的图像存在边缘锯齿化和边缘模糊化缺点。针对该问题,在分析图像复原全变分模型的基础上,将色度-亮度全变分复原模型用于彩色图像放大,使图像放大问题转化为图像修补问题。实验结果证明,该方法能保持放大后图像边缘的光滑与清晰。     

基于Haar小波域自学习的图像超分辨率方法

图像超分辨率重构是指由低分辨率图像来获得高分辨率图像的过程.为了能够有效的重构出高分辨率图像,提出一种基于Haar小波域自学习的图像超分辨率重构算法.该算法将高分辨率图像通过Haar小波变换后得到的近似子块L与已知的低分辨率图像联系起来,然后通过Bp神经网络来自学习Haar小波变换细节子块之间相近的自相似性,从而预测出高分辨率图像通过Haar小波变换后的三个细节子块H,V和D.最后由逆Haar小波变换重构高分辨率图像.实验表明由该算法重构的高分辨图像有很好的视觉效果和峰值信噪比.     

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%。     

A chaos-based symmetric image encryption scheme using a bit-level permutation

In recent years, a variety of chaos-based digital image encryption algorithms have been suggested. Most of these algorithms implement permutations and diffusions at the pixel level by considering the pixel as the smallest (atomic) element of an image. In fact, a permutation at the bit level not only changes the position of the pixel but also alters its value. Here we propose an image cryptosystem employing the Arnold cat map for bit-level permutation and the logistic map for diffusion. Simulations have been carried out and analyzed in detail, demonstrating the superior security and high efficiency of our cryptosystem.