FilterAndColor

神经网络中的对象传播神经网络CPN的学习算法(Neural network objects CPN-propagation neural network learning algorithm)

DISCRETE FRACTIONAL FOURIER TRANSFORM

快速傅里叶变换在Matlab环境下的实现(Fast Fourier Transform in Matlab environment to achieve)

从0开始移植UCOS II到STM32开发板，十分详细的资料，讲的很清楚。手把手教学(From 0 to the UCOS II to STM32 development board, very detailed information, very clear. Hand handle teaching)

通用路径损耗模型包括Log-distance/normal shadowing计算(General Path Loss Model, including log-distance/normal shadowing path loss model)

构建了简单的忆阻神经网络的MATLAB模型，并设计了自适应控制器，来实现驱动-响应同步。(Construct the MATLAB model simple memristive neural network, and the adaptive controller is designed to achieve drive-response synchronization.)

说明：  ％极大似然方法 gcn为高斯有色噪声！！！比较不同点( Maximum likelihood method for Gaussian colored noise GCN! ! ! Compare different points)

一种基于红外图像分层处理及动态压缩的 DDE 算法。该算法先将原始14bits 红外图像数据信息中的大动态低频背景和小动态高频细节进行分离提取，并分别对提取的细节层和背景层进行相应的灰度增强和灰度抑制处理，再调整和压缩各图层的动态范围并最终合成 8bits图 像。实验结果表明，该算法能较好地保留并突出原始红外图像中的边缘和细节信息，达到了预期设计的目标。 (put forward a DDE algorithm based on hierarchical processing and dynamic compression of infrared image． At first，this algorithm dissociates the background ( with wide dynamic and low frequency) and the detail ( with narrow dynamic and high frequency) in the data information ( 14 bits) of original infrared image，and pick up them respectively． Then，it enhances the gray of details layer and restrains the gray of background layer，and readjusts and compresses dynamic range of each layer． Finally，it composites output image ( 8 bits) ． The experimental results show that the proposed algorithm can retain detail information and edges in original infrared image and carry the purpose of design． )

StdAfx.h for builder a C++ implementation.

利用AF-BW filter及双线性变换法设计一DF，满足Wp=0.2p, Ws=0.6p, Ap£ 2dB, As³ 15dB (use AF-BW and bilinear transform filter design a DF, meet Wp = 0.2p, Ws = 0.6p, Ap)