Matlabfreescale

P. Perona&J. Malik算法的改进算法，以及包括了源图片(P. Perona & J. Malik algorithm algorithm, and includes the source image)

1、观察采样引起的混叠。 2、判别离散时间系统的时不变性。 3、利用卷积计算信号通过FIR滤波器的输出，并观察输出信号的input-on暂态、input-off暂态和稳态阶段。 (One to observe the sampling due to aliasing. 2, determine the time invariant discrete-time systems. 3, the use of convolution calculation of the signal through the FIR filter output, and observe the output signal of the input-on transient, input-off transient and steady-state phase.)

说明：  多尺度一维小波分解,用于信号的分解,小波基可以根据需要进行替换.(Multiscale a peacekeeping wavelet decomposition, the decomposition signal for the small Porgy may need replacement.)

基于Kalman滤波器设计思想，仿真实现电池充放电模型建立，实现二阶RC电路充放电模型；(The battery charge and discharge model to achieve second-order RC circuit to charge and discharge model-based Kalman filter design thinking, simulation )

MATLAB7.x数字信号处理 光盘内容。其中为课本的源码。有利于学习。(Digital Signal Processing MATLAB7.x CD-ROM content. One source for textbooks. Conducive to learning.)

学校做的项目源码，运用粒子群算法求解矩阵特征值。粒子群算法注释比较详细，想研究粒子群的同学可以借鉴看看。(Schools do the project source code, the use of particle swarm algorithm for solving matrix eigenvalues. Particle Swarm comment in more detail, the students want to study particle swarm can learn to see.)

说明：  如何使用遗传算法，里面有ppt教程与案例分析，遗传算法与优化问题(How to Use Genetic Algorithms)

比较最佳平方逼近二次多项式，Chebyshev截断级数法以及插值余项极小法这三种函数逼近方法，并给出图形(Comparison of the best square quadratic polynomial approximation, Chebyshev truncated series method as well as more than a very small interpolation method of the three function approximation methods, and gives graphical)

说明：  THIS IS A CODE ABOUT MATLAB

说明：  文件传输限制分多次上传，其他文件可以查看我的上传资料 本实验使用matlab神经网络工具箱搭建CNN网络实现手写字识别，里面附数据集和训练好的网络，准确率可达99%。 1.打开matlab2018 2.将路径设置为当前文件所在路径 3.直接运行test7_run.m即可得到网络在测试集里的准确率以及分类错误的样本。 test7_cnn_hwr.m：训练的程序 test7_run.m：测试测试集的代码，直接读取存储在net_trained.mat训练好的网络， 对测试样本进行分类，统计准确率。(In this experiment, we use matlab neural network toolbox to build CNN network to realize handwritten word recognition, which is attached with data set and trained network, with an accuracy of 99%.)