Survillance_system

meanshift源代码---matlab版，初学者必看(meanshift source code--- matlab version, beginners must-see)

关于matlab基础操作、矩阵运算、程序设计、绘图等的学习资料(Learning materials of the underlying operating on matlab, matrix operations, program design, graphics, etc.)

compressive sensing theory

说明：  matlab实现一些基础的模式识别工作，如贝叶斯分类，聚类算法，bp神经网络(Matlab implements some basic pattern recognition work, such as Bayesian classification, clustering algorithm, BP neural network)

说明：  RBF神经网络控制教材以及书中内含代码例程详解(RBF neural network control textbook and the detailed explanation of the code routines in the book)

MUSIC算法是一种基于矩阵特征空间分解的方法。从几何角度讲，信号处理的观测空间可以分解为信号子空间和噪声子空间，显然这两个空间是正交的。信号子空间由阵列接收到的数据协方差矩阵中与信号对应的特征向量组成，噪声子空间则由协方差矩阵中所有最小特征值（噪声方差）对应的特征向量组成。MUSIC算法就是利用这两个互补空间之间的正交特性来估计空间信号的方位。噪声子空间的所有向量被用来构造谱，所有空间方位谱中的峰值位置对应信号的来波方位。MUSIC算法大大提高了测向分辨率，同时适应于任意形状的天线阵列，但是原型MUSIC算法要求来波信号是不相干的。(MUSIC algorithm is a matrix decomposition method based on feature space. From the geometric point of view, the observed spatial signal processing can be decomposed into signal subspace and noise subspace, it is clear that the two spaces are orthogonal. The signal received by the array to the subspace of the covariance matrix of the signal component corresponding eigenvectors, the noise subspace of the covariance matrix of all by the smallest eigenvalue (noise variance) eigenvectors corresponding to the composition. MUSIC algorithm is the use of orthogonal properties between these two complementary space to estimate the spatial orientation of the signal. Noise subspace of all vectors are used to construct the spectrum, all the spatial orientation of the spectrum corresponding to the peak position of the signal wave direction. MUSIC algorithm greatly improves the resolution measurements, while the antenna array adapted to any shape, but the prototype MUSIC algorithm requires to wave sign)

快速NCC(normalized cross correlation)算法实现。比matlab自带的normxcorr2 快10-20倍。在windows Unix环境下c语言，和matlab编译通过。(use Matlab on Windows or Linux or Mac OS X and want fast (exact, general-case) normalized correlation (NCC) code right now, then download it and enjoy the large performance gain over Matlab s normxcorr2 (demo included). For the brave (and/or masochistic) there is B), however, it is unlikely you will do better than the above. Thanks to Liang Jin for making the OS X binary (his performance comparison graph on a PowerPC). )

matlab中的FFT使用范例，非常经典(FFT in matlab usage examples, very classic)

iRobotCreateSimulatorToolbox1_3 iRobot仿真matlab工具箱(iRobotCreateSimulatorToolbox1_3 iRobot simulation matlab toolbox)

该段程序描述了如何去估计多普勒频率变化率。参考文献为Morelli的《Doppler-rate estimation for digital burst transmission》( This section describes how to program the estimated Doppler frequency rate of change. References to Morelli s "Doppler-rate estimation for digital burst transmission")