please give the coading explanation for the topic drowssiness detection of pilot sir/mam,

Question

Revathi Re 2015-3-29

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/195108-please-give-the-coading-explanation-for-the-topic-drowssiness-detection-of-pilot-sir-mam

回答： balandong 2017-4-10

NEWFINAL.m

Respected sir/mam

I am in need of the coading explanation for this topic within this 31/3/2015

If you replied me on date means it will be a great help for me.

clc close all clear all %% svm Training & Classification

load Train_data; % load group; svmStruct = svmtrain(Train_data,group,'kernel_function','rbf');

%% signal Loading load data; fs = 300; for j=1:3 d=data(j+1,1:6000); x=data1(j+1,1:6000);

%% Pre-processing using LMS mu = 8e-9; ha = adaptfilt.lms(64,mu); [y,e] = filter(ha,d,x); figure,subplot(4,1,1), plot(x),title('Original Signal'); subplot(4,1,2),plot(d),title('Noise with original Signal'); subplot(4,1,3),plot(y),title('Noise Signal'); subplot(4,1,4),plot(e),title('Retrived Original Signal');

%% for i=1:1:length(e)/(2*fs) E=e(1,(i-1)*2*fs+1:i*2*fs); figure, windowsize = 128; window = hamming(windowsize); nfft = windowsize; noverlap = windowsize-1; wpt = wpdec(E,3,'sym8'); [Spec,Time,Freq] = wpspectrum(wpt,fs,'plot'); figure, Hs = spectrum.welch('hamming',26,4);

psd(Hs,Spec,'Fs',fs,'NFFT',512); [Pxx,f] = pwelch(Spec,window,noverlap,nfft,fs); y = moving(Pxx,3,'median'); figure,plot(y) %% Feature collection fea=mean(y(2:end-1)); % fea(2)=std(y); fea1(j,i)=fea;

% fea(3)=kurtosis(y); a1=0; a2=1.1276e-08; b1=8.4479e-04; b2=0.03; if fea<0 F=0; elseif 0<=fea<1.1276e-08 F=0.5*(fea-a1)/(b1-a1); elseif 1.1276e-08<fea< 8.4479e-04 F=0.5; elseif 8.4479e-04<fea<0.03 F=0.5*(fea-a2)/(b2-a2); else F=1; end Fea(j,i)=F; close all end

end

for k=1:size(fea1,2) test_fea=Fea(:,k); %% Testing species(k) = svmclassify(svmStruct,test_fea');

if (Species(k)>0.5) msgbox('The Pilot going to sleep'); else msgbox('The Pilot is Active'); end pause(3); end

With Regards Revathi