wcdma

gaussian process regression for signal processing

特征点提取，一个关于harris算子提取图像特征的程序。。 做了些许改动(Feature point extracting，Harris operator on image feature extraction process. . Make a few changes )

heat 2D code for finite element, very important

ipiq法的谐波检测，谐波源是三相全桥整流电路(ipiq method of harmonic detection, harmonic source is a three-phase full-bridge rectifier circuit)

提出了一种基于小波变换可抗击几何攻击算法。在水印嵌人 部分# 首先对宿主图像进行小波分解 、 嵌人水印, 再通过余弦变换求得该水印图像的一个特征向量, 利用该特征向量和水印 信息通过12 31 函数生成一个二值逻辑系列, 将该序列作为密钥/ 水印提取部分为# 对待测的图像进行小波变换, 再通过余 弦变换求得待测图像的一个特征向量, 利用该特征向量和密钥通过1 2 4 1 函数提取出水印。(Proposed based on wavelet transform algorithm can be against geometric attacks.# Watermark embedded in the first part of the host image wavelet decomposition, embedded watermark, and then through the cosine transform of the watermarked image obtained by a feature vector, using the feature vector and the watermark information through 1231 to generate a binary logic function series, the sequence as the key/watermark extraction part is# treated transform measured image, and then obtained through the cosine transform the image of a feature vector under test, using the feature vector and the key function by 1241 to extract the watermark .)

eigen value decomposition method for acoustic source localization Matlab source codes

一个DataMatrix，RS码的编码例子，用matlab写的，有助于理解里面的算法(A DataMatrix, RS code examples, written with MATLAB, helps to understand the inside algorithm)

分数阶傅里叶变换(Fractional Fourier Transform, FrFT)，实现线性调频(LFM)信号的参数估计(调频斜率和初始频率)。由粗到细搜索分数阶傅里叶变换的阶数。需要到网上自行下载一个工具函数frft.m.(It is about estimationg the frequency modulation rate and initial frequency of linear frequency modulation (LFM) signal using the fractional Fourier transform (FrFT).It need a function frft.m which could be found in the internet.)

在matlab环境下的有关RLS算法,多麦克风语音降噪(In the matlab environment of the RLS algorithm, multi-microphone noise reduction Voice)

小波神经网络的分类识别，学习率因子，网络动量项因子，网络多分辨级数，网络平移参数 都可自己选。本例输入节点5，输出5。可识别五种信号。(wavelet neural network classification, the learning rate factor, network momentum contribution factor, network multiresolution series. Translation network parameters can be of his own choosing. The five cases input nodes, output 5. Five identifiable signal.)