Guided image filtering using data mining and decomposition

Owing to local linear model, guided image filter may contain from ringing artefacts. A modified guided image filter based on data mining techniques and decomposition is proposed in this paper. Image is split into patches, passing through the guided filter, clustered into number of clusters and transformed into compact domain using principal component analysis. Edges are preserved and halo artefacts are removed using detail enhancement based on weighted least square. Simulation results demonstrate the strength of the proposed technique as compared with other techniques especially around the edges where noise and halo artefacts may be present in the image.     

基于复小波变换与层间模型的图像去噪

很多小波去噪方法认为小波系数是相互独立的,然而大量实验表明实际图像的小波系数之间是有较强的依赖性。在本文中,我们将利用复小波变换的优势以及小波系数之间的依赖性,提出一种新的图像去噪方法。该方法先确定滤波器系数,再对复小波变换系数建模,并根据MAP准则给出系数的收缩方法进行去噪处理,最后作复小波逆变换。同时在变换的系数抽取之前估计系数的方差,可以使方差估计更准确。     

Wavelet array decomposition of images using a Hermite sieve

Generalized Hermite polynomials are used in a novel way to arrive at a multi-layered representation of images. This representation, which is centred on the creation of a new class ofwavelet arrays, is (i) distinct from what we find in the current literature, (ii) stable, and (iii) in the manner of standard transforms, transforms the image, explicitly, into matrices of coefficients, reminiscent of Fourier series,but at various scales, controlled by ascale parameter. Among the other properties of the wavelet arrays, (a) the shape of the resolution cell in the ‘phase-space’ is variable even at a specified scale, depending on the nature of the signal under consideration; and (b) a systematic procedure is given for extracting the zero-crossings from the coefficients at various scales. This representation has been successfully applied to both synthetic and natural images, including textures.     

一种采用多个小波基的图像融合去噪方法

提出了一种新的基于多个小波基的图像融合去噪方法.首先利用多个不同的小波基对含噪图像进行阈值去噪,得到多幅恢复图像.然后对这些图像采用小波融合方法进行融合.对于低频系数采用基于边缘的融合算法,在多幅恢复图像中选择最有可能是边缘的点加以保留;对于高频系数,采用了平均的融合算法.最后得到一幅去噪图像.实验结果表明,无论是在视觉效果上还是在峰值信噪比定量指标上该方法去噪效果均明显优于单一小波基去噪.     

基于小波和奇异值分解的图像边缘检测

针对传统图像边缘检测抑制噪声能力弱的问题,给出了一种小波变换和局部梯度场内奇异值分解相结合的边缘检测方法.首先在图像预处理阶段,为了提取准确的边缘特征,文中利用小波变换的时频局部化特性,对图像进行小波变换.该文对用小波求取的梯度场使用局部梯度奇异值分解的方法;利用奇异值的特性和良好的稳定性,使提取的边缘特征更加突出并且能够达到抑制噪声的目的.实验证明该文方法既能在无噪声影响的图像中提取出清晰完整的单边缘,又能在有噪声干扰的情况下提取出理想的边缘.     

一种结合小波变换和维纳滤波的图像去噪算法

目的为了有效消除噪声图像中的椒盐噪声、高斯噪声甚至混合噪声,结合维纳滤波的优势和小波分解各分量的特点,提出一种新的图像去噪算法。方法该算法先将含噪声图像进行小波变换,分离出1个低频分量和3个中高频分量,然后对低频分量进行自适应维纳滤波,对3个中高频分量用Canny算子提取边缘,最后将处理后的4个分量进行重构得到去噪后的图像。结果仿真结果表明,该算法对扫描仪引入的常见噪声均表现出较好的去噪效果,PSNR值均大于20 d B。尤其是对于高斯噪声和混合噪声,新算法去噪后的PSNR结果高于维纳滤波、软阈值小波滤波和文献[9]算法1~8 d B,效果较好。结论结合小波变换和维纳滤波的图像去噪算法,能够较好去除噪声图像的多种类型噪声,是一种较为优秀的去噪算法。     

基于阈值与人眼特性的小波图像压缩算法

针对图像传输数据量大的问题,通过结合小波变换的高效性与普遍性,提出了一种基于阈值与人眼特性的小波图像压缩算法,在图像的低频子带中通过结合阈值与人眼的视觉敏感度,提出了阈值保留算法;而图像的高频子带则是通过结合小波变换后各子带的方向性,并根据小波幅角的不同提出了边缘保留算法：在保证图像质量的同时减少了图像中的冗余信号。实验结果表明,新方法能够实现上述目标且具有低耗时的优势,在图像压缩领域具有较好的发展前景。     

基于双正交小波变换的资源二号卫星与多光谱影像融合方法研究

针对基于Darbechies等正交小波函数的遥感影像融合时引起失真的现象,引入双正交小波函数,提出了基于双正交小波变换的融合方法,对资源二号卫星和TM的多光谱影像的融合进行研究,并把融合结果与主成分、IHS和Brovey等融合结果进行比较。通过对光谱特征曲线、相关系数、光谱扭曲程度等分析和空间细节的对比,说明双正交小波融合方法除了提高空间分辨率外,最好地保持了多光谱影像的光谱特征,有利于专题信息的提取。     

Robust Image Hashing via Random Gabor Filtering and DWT

Image hashing is a useful multimedia technology for many applications, such as image authentication, image retrieval, image copy detection and image forensics. In this paper, we propose a robust image hashing based on random Gabor filtering and discrete wavelet transform (DWT). Specifically, robust and secure image features are first extracted from the normalized image by Gabor filtering and a chaotic map called Skew tent map, and then are compressed via a single-level 2-D DWT. Image hash is finally obtained by concatenating DWT coefficients in the LL sub-band. Many experiments with open image datasets are carried out and the results illustrate that our hashing is robust, discriminative and secure. Receiver operating characteristic (ROC) curve comparisons show that our hashing is better than some popular image hashing algorithms in classification performance between robustness and discrimination.     

基于小波变换与形态学的车牌定位方法

针对信号噪声环境下的车牌定位问题,提出了一种基于小波变换和数学形态学的图像定位方法.该方法先用自适应滤波多尺度边缘检测方法检测出车牌的边缘;然后用数学形态学方法对阈值二值化后的边缘图像进行系列形态运算,进一步消除无用信息;最后用基于车牌底色识别的方法进行车牌定位.实验结果表明,该方法定位效果好,适于有噪声的车牌图像进行定位.     

Efficient Landsat image fusion using fuzzy and stationary discrete wavelet transform

Image fusion is the concept to integrate multiple same scene images while drawing out maximum radiometric information from them by avoiding noise and fictional data. The main objective is to improve the radiometric quality of fused image compared to individual images of the same scene. Existing methods are found to be efficient, but if the similar radiometric information is fused into every image, it produces redundant high frequency of pixels. Therefore, to overcome this issue, in this paper a fuzzy and stationary discrete wavelet transform (FSDWT)-based image fusion technique is proposed. It decomposes Landsat image into stationary values, and then it preserves the radiometric data by using fuzzy if-then rules. In the last phase, FSDWT injects high-frequency blocks from input images and returns a single Landsat image with maximum radiometric data. Quantitative analysis has clearly demonstrated that FSDWT has better structural detail, spatial resolution and spectral information than existing methods.     

一种利用离散余弦变换频谱图和小波变换图像编码的语音编码算法

语音、图像编码技术是现代数字通信的核心技术,通常是分开进行的,提出了一种利用二维DCT频谱图和小波变换图像编码的语音编码算法.深入分析了算法实现的关键技术——语音信号的二维DCT频谱图,仿真实现了该编码算法,对一段语音的处理结果表明,小波分解系数置零率为83.1941％,压缩后保存能量92.0867％,合成语音达到“优...