Low pass filter using filter coefficient
显示 更早的评论
Hi,
I want to do low pass filtering of ECG signal. I have filter coefficients and sammpling frequency. With only this much information available is it possible to do low pass filtering?
please help if someone knows!
Thank you in advance.
采纳的回答
Star Strider
2021-2-16
0 个投票
17 个评论
Fatema Dalal
2021-2-16
Star Strider
2021-2-16
Upload the filter coefficients, and I will see what I can do.
Fatema Dalal
2021-2-16
Star Strider
2021-2-16
I assume these are FIR filters.
If so, implement them similarly to the way I implemented them in:
Fs = 1000; % Use The Correct Sampling Frequency
LPF_b = [ 0.0352 -0.0854 -0.1350 0.4599 0.8069 0.3327];
HPF_b = [-0.3327 0.8069 -0.4599 -0.1350 0.0854 0.0352];
figure
freqz(LPF_b,1, 2^14, Fs) % Filter Bode Plot
title(subplot(2,1,1), 'Lowpass FIR Filter')
figure
freqz(HPF_b,1, 2^14, Fs) % Filter Bode Plot
title(subplot(2,1,1), 'Highpass FIR Filter')
EKG_data = rand(1, 1000);
If you want to filter the signal first with the lowpass filter and second with the highpass filter, the filter arguments would be:
LPF_EKG_data = filtfilt(LPF_b,1, EKG_data); % Filter Using Lowpass Filter
HPF_EKG_data = filtfilt(HPF_b,1, LPF_EKG_data); % Filter Using Highpass Filter
With the result of both filters being the output of the highpass filter.
Experiment to get the result you want.
See the documentation for the various funcitons for details on how to use them.
Fatema Dalal
2021-2-23
Star Strider
2021-2-23
As always, my pleasure!
Fatema Dalal
2021-3-14
Star Strider
2021-3-14
A bandpass filter is most appropriate for that, since it will remove baseline variations as well as high-frequency noise. EKG signals have a bandwidth of 0 to 100 Hz, so the filter passband should be 1 Hz to 100 Hz, although for a normal EKG (without arrhythmias or significant variations in the S-T segment), a filter with a passband of 1 to 45 Hz will work. That will also filter out mains frequency (powerline) noise.
Fatema Dalal
2021-3-15
Star Strider
2021-3-15
The easiest way to do that is to let the bandpass function design it. The digital filter object (with all the necessary information) is the second output of the function. All you need to do is to decide what the upper passband frequency should be, and use 1 Hz for the lower passband frequency.
Fatema Dalal
2021-3-16
Star Strider
2021-3-16
If you have the Signal Processing Toolbox, the bandpass function (and related functions) do all the coding for you. They were introduced in R2018a, so if you have an earlier version I can provide you with appropriate code using an elliptic filter. It would be best to try the bandpass function first.
If you do not have the Signal Processing Toolbox, filtering anything is going to be difficult, although not impossible.
Star Strider
2021-3-16
Fatema Dalal’s Aswer is now this Comment —
I do have the singnal processing toolbox. But I dont know how to use it.
Star Strider
2021-3-16
If you have R2018a or later version of it, please read the documentation for the bandpass function that I linked to earlier. It will tell you everything you need to know about how to use the function. If you have an earlier version, I can post equivalent code to design an elliptic filter and filter the signal with it.
Fatema Dalal
2021-3-16
Star Strider
2021-3-16
The easiest way to determine the locations (and amplitudes) of the R-deflections is to use the findpeaks function.
Fatema Dalal
2021-3-16
更多回答（0 个）
类别
在 帮助中心 和 File Exchange 中查找有关 Single-Rate Filters 的更多信息
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!
Translated by 
