本课题为基于MATLAB的小波变换dwt和离散余弦dct的多方法对比数字水印系统。带GUI交互界面。有一个主界面GUI，可以调用dwt方法的子界面和dct方法的子界面。流程包括，读取宿主图像和水印图像，嵌入，多种方法的攻击（剪切，加噪，旋转等），提取，最后利用psnr峰值信噪比进行评价不同攻击下，鲁棒性的好坏。
This project is based on MATLAB wavelet transform DWT and discrete cosine DCT multi method contrast digital watermarking system. With GUI interactive interface. There is a main interface GUI, which can call the sub interfaces of DWT method and DCT method. The process includes: reading host image and watermark image, embedding, attacks of various methods (cutting, adding noise, rotation, etc.), extracting. Finally, PSNR peak signal-to-noise ratio is used to evaluate the robustness of different attacks. (2020-08-07, matlab, 9766KB, 下载27次)
该算法首先通过字典学习得到含噪图像的冗余字典，然后对相似的图像块进行聚类构成块群，并通过迭代收缩和L1正则化约束，对同类的图像块在字典上进行稀疏表示，以达到降噪的目的。实验结果表明，在常规的图像处理上，本文提出的算法能较好的保留图像的结构信息，与K-SVD和BM3D等现有的流行算法相比，具有更高的峰值信噪比（PSNR）
It firstly get the redundant dictionary of a noised image by dictionary learning.Then,the image patches are gathered according to their similarities.Meanwhile,the similar patches get sparse representation showed in dictionaries by iterative shrinkage and L1 regularization constraints and eventually the image is restored and noise is removed.The experimental results indicate that the proposed algorithm can well preserve the structure information of the common image with a higher Peak Signal to Noise Ratio(PNSR),compared with state-of-the-art algorithms,such as K-SVD and BM3D (2017-05-02, matlab, 1810KB, 下载37次)
基于小波多分辨率分析的思想，提出一种降低混沌信号噪声的方法。首先对一个混沌时间序列进行小波分析，得到信
号在各个尺度上的不同特性 再对各个高频细节信号进行二次小波去噪 最后对二次小波去噪后的信号进行线性叠加，达到
提取真实信号的目的。分别对Rossler 系统产生的混沌序列和黄河年径流量混沌序列进行了仿真研究，结果表明了该方法的
有效性。
A method was proposed to reduce noise within chaotic signal based on mulrti-resolution wavelet transform.
Based on the wavelet analysis ，different characters of chaotic signal were observed on different scales.
Then ,each detail with high frequencies was denoised by wavelet for the second time in order to observe the real
signal. Chaotic time series generated by Rossler systemand the practically observed yearly runoff of Yellow River
were respect ively applied for noise reduct ion. Simulat ion results show that this method is feasible to noisy chaot ic
signal. (2017-01-19, matlab, 279KB, 下载5次)
在心电信号处理过程中，为了避免产生Gibbs 振荡现象和严重的频率混叠现象，提出一种基于双树复小波变换，并结合最大后验估计确定阈值的心电信号去噪方法。文中采用了信噪比和均方误差来评价双树复小波变换和离散小波变换两种方法对心电信号的去噪效果。实验结果表明: 与传统离散小波变换相比，双树复小波变换去噪更彻底，边界、纹理等特征能较好地保留，可以作为一种生物医学信号降噪处理的新方法。
In ECG signal processing， in order to avoid the phenomena of Gibbs oscillation and severe frequency aliasing，a new ECG signal denoising algorithm is presented，which is based on dual-tree complex wavelet transform and combined with the maximum a posteriori estimation to determine the threshold． The signal-to-noise ratio and mean square error are used to uate the denoising effects of the dual-tree complex wavelet transform and discrete wavelet transform． The experimental result shows that compared with traditional discrete wavelet transform， the dual-tree complex wavelet transform reduces noise more thoroughly and retains boundary and texture characteristics better． The dualtree complex wavelet transform can be used as a new denoising method for biomedical signal denoising processing． (2016-12-23, matlab, 854KB, 下载20次)
本论文包含5个源程序，分别为LSBCut、LSBCollage、DCTCut、DCTCollage和PSNR，其中LSBCut实现了LSB水印算法的剪切篡改的检测、LSBCollage实现了LSB水印算法的拼贴篡改的检测、DCTCut实现了DCT水印算法的剪切篡改的检测、DCTCollage实现了DCT水印算法的拼贴篡改的检测、PSNR功能是计算嵌入水印嵌入的峰值信噪比，以此来说明嵌入水印后图像载体的失真大小。LSBCut、LSBCollage、DCTCut、DCTCollage是独立的程序，每个程序独立完成一种篡改检测的功能且都有一个输出参数，没有输入参数，输出的参数是漏警概率。PSNR需要LSBCut、LSBCollage、DCTCut、DCTCollage产生的嵌入水印后的图像和原始图像作为输入参数，作为LSBCut、LSBCollage、DCTCut和DCTCollage的内函数使用。
This thesis contains five source program, LSBCut, LSBCollage respectively, DCTCut, DCTCollage and PSNR, including LSBCut implements the LSB watermarking algorithm of shear tampering detection, LSB watermarking algorithm was realized LSBCollage collage tampering detection, DCTCut implements the DCT watermarking algorithm of shear tampering detection, DCTCollage implements the DCT watermarking algorithm of collage of tampering detection, PSNR function is to embed watermark embedding peak signal-to-noise ratio, in order to show that embedded watermark image carrier after distortion.LSBCut, LSBCollage, DCTCut, DCTCollage is independent of the programs, each program independently a tamper detection function and has an output parameter, no input parameters, output parameters are leakage alarm probability.PSNR need LSBCut, LSBCollage, DCTCut, DCTCollage of embedded watermark image and the original image as the input parameters, as LSBCut, LSBCollage, DCTCut and DCTCollage function in use. (2016-10-24, matlab, 19940KB, 下载27次)
图像处理（图像融合，图像加密等）的性能评价指标集包括：平均梯度，边缘强度，信息熵，灰度均值，标准差(均方差MSE)，均方根误差，峰值信噪比（psnr），空间频率(sf)，图像清晰度，互信息(mi)，结构相似性(ssim)，交叉熵(cross entropy)，相对标准差。
Image processing，average gradient, edge strength, information entropy, gray mean, standard deviation (variance (MSE), are root mean square error, peak signal to noise ratio (PSNR), spatial frequency (SF), the sharpness of the image, the mutual information (MI), similar to the structure of (SSIM) and cross entropy (cross entropy). The relative standard deviation (2016-07-12, matlab, 8KB, 下载75次)
纺织品制造商必须监控其产品的质量，以保持既定用于服装行业的高品质标准。因此，纺织品的质量控制是为他们的企业的竞争力的增加的关键因素。纺织断层，传统上被人类目测检测。然而，人的检查是耗时的，并且不实现精度较高的水平。因此，工业视觉单元是用于纺织工业的战略利益，因为他们可以组成一个系统实现高精度纺织检测高度的基础。这项工作说明专为简单的图像处理操作的基础上，验布，以及其在检测到通常的纺织品缺陷效率系统的软件核心。整个系统的先决条件是，然后解析 所讨论的，以及由于该问题的本质上的限制和限制。软件算法和第一结果的评价也列于细节。
Textile manufacturers have to monitor the quality of their products in order to maintain the high-quality standards established for the clothing industry. Thus, textile quality control is a key factor for the increase of competitiveness of their companies. Textile faults have traditionally been detected by human visual inspection. However, human inspection is time consuming and does not achieve a high level of accuracy. Therefore, industrial vision units are of strategic interest for the textile industry as they could form the basis of a system achieving a high degree of accuracy on textile inspection. This work describes the software core of a system designed for fabric inspection on the basis of simple image-processing operations as well as its effi ciency on detection of usual textile defects. The prerequisites of the overall system are then discussed analytically, as well as the limitations and the restrictions imposed due to the nature of the problem. The software algorithm a (2014-05-26, matlab, 416KB, 下载13次)
针对传统的空间域纹理图像修复算法计算量大、修复时间长的缺点, 本文提出了一种纹理图像的快速修复算法。该算
法的基本思想是在小波域中利用小波系数的能量来确定待修复块的填充顺序, 并结合纹理合成的方法填充待修复区。实验结
果表明, 该算法不仅可以大大提高纹理图像的修复速度, 而且在峰值信噪比和主观视觉效果上都优于传统的图像修复算
Trad itional spatia-l doma in inpainting algorithm s for tex ture image need comp licated computat ion and consume
a long t ime. A fast inpainting algorithm for tex ture image is proposed by determ in ing the order o f the filling
blocks using the energ ies o f w avelet coeffic ients in w ave let dom ain, and filling the inpa in ting b locks using tex ture
synthesis method. Experimental resu lts show that the a lgorithm not on ly improves inpainting speed g reat ly, but a lso
achieves better peak signa l no ise rat io ( PSNR ) and higher v isual qua lity.
(2014-03-22, matlab, 791KB, 下载21次)
给出了一种新的基于小波多尺度分解的分层图像融合方法. 其基本思想是先对源图像
进行小波多尺度分解 其次, 按照融合规则, 采用基于区域特性量测的选择及加权算子去构造融
合图像对应的小波系数 最后, 通过逆小波变换重构融合图像. 该方法被成功地用于图像的融合
处理. 此外, 利用熵、交叉熵、互信息、均方根误差、峰值信噪比等参量, 对该融合方法的融合性能
进行了评价与分析. 实验结果表明, 该融合方法是十分有效的
A novel h ierarch ical image fusion scheme based on wavelet mult iscale de2
compo sit ion is p resented. The basic idea is to perfo rm a wavelet mult iscale decompo si2
t ion of each source image first, then the wavelet coefficient s of the fused image are
const ructed using region2based select ion and weigh ted operato rs acco rding to different
fusion rules, finally the fused image is obtained by tak ing inverse wavelet t ransfo rm.
Th is app roach has been successfully used in image fusion. In addit ion, w ith the use of
the parameters such as ent ropy, cro ss ent ropy, mutual info rmat ion, roo t mean square
erro r, peak2to2peak signal2to2no ise rat io, the perfo rmance of the fusion scheme is e2
valuated and analyzed. The experimental result s show that the fusion scheme is effec2
t ive. (2012-10-30, matlab, 376KB, 下载62次)
语音增强算法，谱减法、小波变换以及数学形态学等的语音增强算法针对信噪比、语谱图、去噪情况，分析比较
Speech enhancement algorithm, spectral subtraction, wavelet transform and mathematical morphology speech enhancement algorithm for signal-to-noise ratio, the spectrogram denoising situation analysis (2012-10-22, matlab, 1087KB, 下载338次)
本程序主要用两种方法对前后输入两幅RGB图像计算处理，分别计算了其均方差MSE和峰值信噪比PSNR；方法1:如果读入图像为彩色图象, 首先进行灰度化处理,依照灰度图象计算
方法2:对RGB图像均方差是所有方差之和除以图像尺寸再除以3
This program before and after the input of two ways two RGB image calculation process to calculate the mean square error MSE and PSNR PSNR Method 1: If you read into the image as a color image, the first gray-scale processing Method 2: RGB image variance calculated in accordance with the gray-scale image is all variance divided by image size divided by 3 (2012-10-10, matlab, 6KB, 下载70次)
图像修复是图像处理的一个重要问题。这是一个困难的问题，同时填写在地区缺少图像信息的纹理和结构。以损坏的图像inpaint失踪的结构和纹理信息，基于小波分解的图像修补算法提出。首先损坏的图像分解成子图像和纹理结构，利用小波子图像转变。然后，地区结构中缺少的信息的子图像重建曲率驱动扩散（CDD），算法，而在同一区域的纹理子图像填写基于改进的纹理合成的典范 最后，恢复的图像，通过重组，结构和纹理恢复的结果。大量的实验表明，该算法可以快速，高效地恢复在同一时间的结构和纹理信息，视觉效果和峰值信噪比比（PSNR）比同类算法更好。
Image inpainting is an important problem image processing. It is a difficult problem to simultaneously fill-in the
texture and structure in regions of missing image information. In order to inpaint the damaged image with both
missing the structure and texture information, an image inpainting algorithm based on wavelet decomposition is
presented. First the damaged image is decomposed into structure sub-image and texture sub-image using the wavelet
transformation. Then, the sub-image with the region of missing information in the structure is reconstructed by
Curvature-Driven Diffusions (CDD) algorithm, while the same region in the texture sub-image is filled-in with the
improved texture synthesis based on exemplar Finally, the restored image is given by recombining the structure and
texture restored results. A large number of experiments show that the proposed algorithm can quickly and efficiently
restore the structure and texture information at the same time, and the visual eff (2012-06-05, matlab, 240KB, 下载171次)