descomposition

实现了对10个数字音的识别，是学习PCA特征提取的很好的学习资料，包括调制，解调，信噪比计算。( To achieve the recognition of 10 digital sound, Is a good learning materials to learn PCA feature extraction, Includes the modulation, demodulation, signal to noise ratio calculation.)

贝叶斯小波去噪是基于小波变换得软阈值贝叶斯去噪算法。(Bayesian wavelet denoising is wavelet-based Bayesian denoising was soft-thresholding algorithm.)

这是图象缩放的代码，对于学习图像的朋友应该有所帮助(This is the image scaling code for the study should help the image of a friend)

一种图像配准算法源码，实现了图像的配准拼接，看看就知道好。(A source image registration algorithm to realize image registration splicing to see will know better.)

边缘检测法和阈值分割法，是图像分割技术应用中两项重要的方法。本文介绍了图像边缘检测和阈值分割的一般原理、典型算法以及研究现状。本文针对边缘检测算子和几种阈值分割技术进行了详细讨论。针对火柴根数统计和钢筋根数统计两个实验，在详细分析检测原理的基础上，编写了相应的实验并给出了实验结果。(Edge detection and thresholding method, image segmentation is an important technology in two ways. This article describes the general principles of image edge detection and thresholding, the typical algorithms and research status. In this paper, several edge detection operator and thresholding techniques are discussed in detail. Statistics for the number of matches and reinforced the number two experimental statistics, based on a detailed analysis of the detection principle, write the corresponding experimental and experimental results are given.)

采用曲率的方法，等间隔三点的大小计算夹角的大小，根据阈值标记指尖指根区域。(Curvature of the methods used, such as the size of three intervals calculated the size of the angle, according to the threshold refers to the root area marked fingertips.)

APG算法，加速近端临近点算法，通过对子问题的调整，计算速度很快(APG algorithm, accelerated near the proximal point algorithm, by adjusting the sub-problems, fast calculation)

基于主动学习的高光谱遥感图像分类，主动学习方法也比较流行，新颖。(An Active Learning Approach toHyperspectral Data Classification)

利用ADMM和小波进行图像恢复 使用ADMM算法的图像恢复模型，并用到了增广拉格朗日函数和快速傅里叶函数(Using ADMM and wavelet to restore image, using ADMM algorithm to restore image model, and using augmented Lagrange function and fast Fourier function.)

下面是利用卡尔曼滤波的方法，实现雷达对目标的跟踪:一目标沿水平方向运动，起始位置为（－2000米，1000米），运动速度为15米/秒，扫描周期T＝2秒， 米，采用蒙特卡洛方法对跟踪滤波器进行仿真，仿真次数为100次。(Below is the use of Kalman filtering method, to achieve the target tracking radar. Parameter :: a target in a horizontal direction, the starting position (2000m, 1000m), the motion speed of 15 m/s, the scan period T = 2 seconds, rice, Monte Carlo method using the tracking filter simulation, simulation 100 times.)