How to remove Noise using histogram In Matlab

Hello Everyone, i hope you are doing well. I have the following data in which I have some noise. I want to remove the Noise using Histogram or any other Method. I have attached the picture below which is the noise.

How can i do that?

Answer 1

[Abderrahim. B]

Hi!

Below is a workflow based on groupcounts function:

clear
load noisydata.mat
figure
subplot(211)
plot(dataset, 'o')
title("noisy data")
% Filtering based groupcounts
[GC, GR] = groupcounts(dataset) ;
denoisedData = dataset(ismember(dataset, GR(GC>10))) ; % group that has count less than 10 will be removed
subplot(212)
plot(denoisedData, 'o')
title("denoised data")

Hope this helps

[Stephen john]

@Issa Still there is noise as you can see the following Imaage

[Abderrahim. B]

Hi !

I recommend to use a slider to vary countThresh variable.

clear
load noisydata.mat
figure
subplot(211)
plot(dataset, 'o')
title("noisy data")
% Filtering based groupcounts
[GC, GR] = groupcounts(dataset) ;

countThresh = 20 ; % change this untill you see that the data is fully denoised
denoisedData = dataset(ismember(dataset, GR(GC>countThresh))) ; % group that has count less than 10 will be removed
subplot(212)
plot(denoisedData, 'o')
title("denoised data")

[Stephen john]

@Issa Getting the following Error While using this function

Error using cell/ismember (line 34)
Input A of class double and input B of class cell must be cell arrays of character vectors, unless one is a character vector.

[Abderrahim. B]

@Stephen john show your code. For me working just fine.

[Stephen john]

Here is the Dataset and your code it gives error

denoisedData is empty

[GC, GR] = groupcounts(Bindata) ;
countThresh = 30 ; % change this untill you see that the data is fully denoised
denoisedData = Bindata(ismember(Bindata, GR(GC>countThresh))) ; 

[Abderrahim. B]

Please show up what print executing the below command:

class Bindata

[Stephen john]

@Issa The Bindata is the array which contain data and class Bindata shows 'char'

[Abderrahim. B]

Sorry wanted to write

whos Bindata

Maybe you can upload the data from where you got Bindata!

[Stephen john]

here is the data

[Stephen john]

Name Size Bytes Class Attributes

Bindata 503x1 4024 double

[Abderrahim. B]

Check the attached mlx script also.

clear
load dataset.mat
figure
subplot(211)
plot(Bindata, 'o')
title("noisy data")
ylim([20 40])
xlim([0 600])
% Filtering based groupcounts
[GC, GR] = groupcounts(Bindata) ;
countThresh = 11 ; % change this untill you see that the data is fully denoised
denoisedData = Bindata(ismember(Bindata, GR(GC>countThresh))) ; % group that has count less than 10 will be removed
subplot(212)
plot(denoisedData, 'o')
title("denoised data")
ylim([20 40])
xlim([0 600])

[Stephen john]

@Issa I need a uniform code which works on all datasets like that , In the above code of your I need to change the countThresh = 11 which is not a good way.

[Abderrahim. B]

I understand, try other filtering methods. This method - based find groups count requires to set a threshold for filtering.

[Stephen john]

@Issa Can you help me with this using Histogram?