AMmatlab

背景提取，迭代法(Surendra算法）；即采用帧差的方法，一帧图像减去前一帧图像，即可以对比得到图像中变化（移动）的区域，扣除移动的区域，即得到不动的背景。采用多帧图像迭代的方法，得到更加准确的背景。(Background extraction, iterative method (Surendra algorithm) frame difference method is used, an image by subtracting the previous frame, which can receive the image contrast change (movement) of the area, less the area of ​ ​ movement, that is obtained immovable background. Multi-frame image iteration method, to be more accurate background.)

ICA（主分量分析）算法和程序，阵列信号处理的高分辨率估计，随机调制信号下的模拟ppm，有信道编码，调制，信道估计等，主要为数据分析和统计，模式识别中的bayes判别分析算法，虚拟力的无线传感网络覆盖，有较好的参考价值。( ICA (Principal Component Analysis) algorithm and procedures, High-resolution array signal processing estimates, Random ppm modulated analog signal under Channel coding, modulation, channel estimation, Mainly for data analysis and statistics, Pattern Recognition bayes discriminant analysis algorithm, Virtual power wireless sensor network coverage, There are good reference value.)

用FORTRON程序编的最速下降法，用来解决无约束规划问题(With FORTRON program compiled the steepest descent method, which is used to solve the unconstrained programming problem)

研究用于自适应均衡器的LMS算法。研究步长的影响。分别画出W=2.9时，mu= 0.01、0.04和0.08情况下的MSE学习曲线(Research for the adaptive LMS algorithm equalizer. Research on the impact of the step. Draw W = 2.9, respectively, when, mu = MSE learning curve 0.01,0.04 and 0.08 in case of)

matlab与组态王的DDE连接，每隔一秒将组态王的数据传给matalb，这样组态王的界面设计就可与matlab连接，实现算法运用。(Matlab and configuration king of DDE connection, will configuration king of data to matalb, such configuration king interface design can connect with matlab, using the algorithm. )

基于SPWVD变换的跳频信号参数估计和检测机器应用(SPWVD transform-based frequency-hopping signal parameter estimation and testing machines)

计算MIMO系统中信道容量的程序，包括中断容量和遍历性容量两种(CITIC MIMO system channel capacity calculation procedures, including the outage capacity and ergodic capacity )

说明：  利用BP神经网络设计指纹识别算法，给神经网络提供每一模式类中的一些样本作为训练样本。BP网络经过学习，不仅能够识别已训练过的样本，而且能够识别未出现过的样本。利用神经网络的泛化能力，提高指纹识别算法的识别率，图像增强、提取指纹特征(BP neural network is the first mock exam to design fingerprint recognition algorithm, and it provides samples for each neural network as training samples. After learning, BP network can not only recognize the trained samples, but also recognize the samples that have not appeared. The generalization ability of neural network is used to improve the recognition rate of fingerprint recognition algorithm, image enhancement and fingerprint feature extraction)

code of an egress router for a mail management system

说明：  包括模糊聚类分析数据各种标准化变换、模糊聚类分析建立模糊相似矩阵的各种方法(Fuzzy Cluster Analysis, including a variety of standardized data transformation, fuzzy cluster analysis to establish fuzzy similar matrix of the various methods)