1kernel-ica1_1

迪杰斯特拉算法求最短路径matlab的实现算法(Dijkstra shortest path algorithm)

清华大学信号处理原理课程第一次实验，内容为产生固定波形以及声波的叠加(Signal processing theory, Tsinghua course first experiment, the content and the acoustic waveform to generate a fixed superposition)

说明：  fmincon finds a constrained minimum of a function of several variables. fmincon attempts to solve problems of the form: min F(X) subject to: A*X

说明：  基于小波变换模极大的多尺度图像边缘检测matlab源代码，该算法采用的是样条小波，为了更好的检测边缘，用多孔算法代替了mallat算法。该算法和mallat快速算法主要区别： 1 多孔算法不需要抽取偶数样本，所以奇异检测定位更准确，相应的重构是不需要插零。2 多孔算法需要对滤波器进行伸缩。(based on wavelet transform modulus maxima of multiscale edge detection Matlab source code, The algorithm used is kind of wavelet, in order to better detect the edge of porous algorithm instead of Mallat algorithm. The algorithm and fast algorithm Mallat major difference : an algorithm porous not even take samples So singular position detection more accurate, the corresponding Reconstruction interpolation is not zero. Two porous algorithm need to filter extendable.)

计算超声波入射到介质中，与缺陷作用时，其纵波衍射与横波衍射的衍射系数(When calculating the ultrasonic incident to the media, and defects, its longitudinal wave diffraction and the diffraction coefficient of the shear wave diffraction)

Discrete Hopfield neural network classification

Load Frequency Control using ANN NARMA-L2

基于遗传算法，进而优化惩罚因子，核函数参数的的支持向量机进行分类(GA SVM classification)

说明：  小波在信号处理中的应用，利用小波对信号进行分解和重构，而且对信号进行噪声的阈值处理。(Wavelet in signal processing applications, the use of wavelet decomposition and reconstruction. but for the noise signal processing threshold.)

阵列处理课程作业，最优权波束形成，理论值、和不同快拍的对比。(Array processing course work, the optimal weight beamforming, the theoretical value, and snapshot comparison.)