DTW_speech_recognition

旋转空间不变法，是阵列信号处理中经典方法，高分辨率DOA估计(espirit algorithm is widely used in high resolution DOA estimation)

Packet-dropfunction functionfunction function

实现ECG和BCG信号采集和同步监测，对采集的信号进行了特征量提取并作出相应诊断，波形显示在液晶屏上。(Achieve ECG and BCG signal acquisition and simultaneous monitoring of the collected signals feature extraction and make the appropriate diagnosis, the waveform displayed on the LCD screen.)

最小二乘，极大似然定位算法，其中包括二维和三维的最小均方误差定位算法以及连续定位算法(Least squares, maximum likelihood localization algorithm, which includes two-dimensional and three-dimensional minimum mean square error localization algorithm and continuous positioning algorithm)

通过matlab设计的GUI界面实现语音信号的实时录入和频谱分析和滤波等处理，并能够显示各阶段时域波形及频谱图，滤波器种类及参数可调(Through the GUI interface designed by matlab, the real-time input of speech signal, spectrum analysis and filtering are realized, and the time-domain waveforms and spectrograms of each stage can be displayed, and the types and parameters of filters can be adjusted.)

MATLAB数值计算，matlab实现，与《matlab数值计算》一书配套使用。(MATLAB numerical calculation, matlab realize, with " matlab numerical calculation," a book supporting the use of.)

这是一个非常适合初学者的Matlab教程，里面讲的东西简洁明了，而且可以作为教学用。(This is a tutorial for beginners in Matlab, which was about things concise, and can be used as a teaching.)

PI控制的仿真模型模型，以及参数设置方面的相关问题。(PI control simulation model models, and parameter settings aspects related issues)

心电信号R波峰值提取算法，抗干扰能力非常强。(ECG R-wave peak extraction algorithm, anti-interference ability is very strong.)

说明：  本文设计了一款人体行为异常监控系统，主要适用人群是老年人，在摄像头固定的情况下，自动检测人体运动轨迹，并与提前设定好的行为库进行匹配，分析判断是否具有异常行为。 在数字图像预处理部分采用了图像二值化，腐蚀与膨胀等几种方法为人体目标的跟踪和检测做准备。为了克服在实际操作中遇到的问题，采用了帧差法和ViBe算法，帧差法即利用帧间变化与当前帧、背景算法来判断它是否大于阈值，并分析视频中序列的运动特性，ViBe算法则是一种背景建模的方法，背景模型是由邻域像素来创建，并对比背景模型、当前输入像素值检测出前景，确定视频中的目标跟踪。在人体行为识别中，运动目标最小长宽比以及连续帧间的加速度来判断人体行为是否异常，如果检测到异常的行为比如说摔倒、快跑等行为，在识别的过程这种实时监测。不同环境可能需要调试，要求有一定编程基础，小白慎入。(In this paper, a monitoring system for abnormal human behavior is designed, which is mainly applicable to the elderly. When the camera is fixed, it can automatically detect the human motion track and match with the behavior database set in advance to analyze and judge whether there is abnormal behavior.)