MATLAB Answers

Attempting to detrend and remove noise from surface roughness data

1 view (last 30 days)
Ryan Garcia
Ryan Garcia on 5 Sep 2021
Commented: Image Analyst on 6 Sep 2021
I am currently trying to remove the waviness from my data and also remove the noise from it to get the surface roughness profile out of the data set that I need. I attached what the graph looks like after attempting to use the detrend function on the data, but it still has significant waviness to the data set. Any help would be appreciated. Below is the code that I am currently working with. I attmepted to use a fft to remove the noise, but the data points are not equally spaced.
clc
clear all
close all
%Clear all variables and open figures and clear the command window
dt = 0.001
t = 0:dt:1;
[file,path] = uigetfile(); %Opens the file browser to select the desired file
imported_data = readtable(strcat(path,file)); %Reads .DAT file into a table
x_height = imported_data{:,2}; %Saves the first column of the data to the variable
z_distance = imported_data{:,1}; %Saves the second column of the data to the variable
x_height_corrected = detrend(x_height); %Removes the negative slope of the data
% n = length(z_distance); %finds length of data set
% fhat = fft(x_height_corrected, n); %Computes the fast foureir transform for data set
% PSD = fhat.*conj(fhat)/n; %Power spectrum (power per freq)
% freq = 1/(-1.96848491018642e-05*n)*(0:n);
% L = 1:floor(n/2);
% f = (0:n-1)/n;
% Plots the data that is imported and detrended
plot(z_distance, x_height_corrected)
ylabel('Height (mm)')
xlabel('Distance (mm)')
  1 Comment
Image Analyst
Image Analyst on 6 Sep 2021
Please attach screenshots in PNG form using the frame or paperclip icon. If they are a .fig file, clicking on it launches a whole, new, separate second instance of MATLAB which takes time. If you put in a picture here directly with the frame icon we can avoid all that trouble.

Sign in to comment.

Answers (1)

Mathieu NOE
Mathieu NOE on 6 Sep 2021
hello
my prefered function for smoothing data is smoothdata , but there are other options
example below :
clc
close all
Fs = 1000;
samples = 1000;
dt = 1/Fs;
t = (0:samples-1)*dt;
y = square(2*pi*3*t) + 0.1*randn(size(t));
% %%%%%%%%%%%%%%%%
figure(1)
N = 25;
ys = smoothdata(y, 'gaussian' , N);
plot(t,y,t,ys);legend('Raw','Smoothed');
title(['Data samples at Fs = ' num2str(round(Fs)) ' Hz / Smoothed with smoothdata' ]);
% %%%%%%%%%%%%%%%%
figure(2)
N = 25;
ys = medfilt1(y, N,'truncate');
plot(t,y,t,ys);legend('Raw','Smoothed');
title(['Data samples at Fs = ' num2str(round(Fs)) ' Hz / Smoothed with medfilt1' ]);
grid on
%%%%%%%%%%%%%%%%
figure(3)
ys = sgolayfilt(y,1,21);
plot(t,y,t,ys);legend('Raw','Smoothed');
title(['Data samples at Fs = ' num2str(round(Fs)) ' Hz / Smoothed with sgolayfilt' ]);
grid on
%%%%%%%%%%%%%%%%
NN = 2;
Wn = 0.25;
[B,A] = butter(NN,Wn);
figure(4)
ys = filtfilt(B,A,y);
plot(t,y,t,ys);legend('Raw','Smoothed');
title(['Data samples at Fs = ' num2str(round(Fs)) ' Hz / Smoothed with butterworth LP' ]);
grid on

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!