Omit detrending in certain portion of signal
조회 수: 9 (최근 30일)
이전 댓글 표시
Hello,
I have a signal with a certain portion that is much higher than the rest of the signal.
How can I avoid applying detrending to this portion but apply detrending to everything to the left and right of this signal? Blue portion should be detrended. Red portion should not be detrended and stay at this or a similar level.
What if I have two of these red portions within my signal? For example the lower peak portion in the middle and the higher portion to the right. Do not detrend both portions?
Thanks.
댓글 수: 2
채택된 답변
Mathieu NOE
2022년 3월 23일
hello
may I suggest a small improvement here - this avoid finding manually the samples that shoudn't be detrended
so whatever the amount and x position of the peaks, this code should work :
load('signal.mat');
n = 1:length(signal);
n = n';
signalS = smoothdata(signal,'movmedian',200);
I = abs(signal-signalS) < 5e-3;
signal_detrended = detrend(signal(I), 'SamplePoints', n(I));
signal_combined = zeros(length(signal), 1);
signal_combined(I) = signal_detrended;
signal_combined(~I) = signal(~I);
figure(1), plot(n, signal,n, signalS,n, signal_combined)
legend('raw','smoothed','partial detrend');
추가 답변 (1개)
Benjamin Thompson
2022년 3월 22일
You can use an index vector and the 'SamplePoints' option for the detrend function:
>> n = 1:length(signal);
>>n = n';
>> I = (n < 4000) | ((n > 4500)) & (n < 8500);
>> signal_detrended = detrend(signal(I), 'SamplePoints', n(I));
>> hold off;
>> figure, plot(n, signal)
>> grid on, zoom on
>> hold on
>> plot(n(I), signal_detrended, 'r')
댓글 수: 2
Benjamin Thompson
2022년 3월 22일
This is just some basic MATLAB commands, but here you are. The index vector defines the set of what was detrended and what was not.
>> signal_combined = zeros(length(signal), 1);
>> signal_combined(I) = signal_detrended;
>> signal_combined(~I) = signal(~I);
>> figure, plot(n, signal_combined)
참고 항목
카테고리
Help Center 및 File Exchange에서 Spectral Measurements에 대해 자세히 알아보기
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!