Help Segmenting signal processing

Hi,

I have data. I have segmented the signal in windows and calculated rms. Can you please help in finding multiple minimum values for this signal?

[v, i] = min(rms);

Thank you!

Answer 1

[Mathieu NOE]

hello

see my little demo below. I assumed that you would split the rms data in 3 groups corresponding to ranges

• 0 to 33% of max rms
• 33 to 66% of max rms
• above 66% of max rms

the 3 groups are the colored dots

code :

clearvars
% dummy data
n=300;
x=linspace(0,2*pi,n);
data = 0.25*ones(size(x));
data = max(data,-cos(x));

buffer = 10;  % nb of samples in one buffer (buffer size) 
overlap = 5; % overlap expressed in samples

%%%% main loop %%%%
m = length(data);
shift = buffer-overlap;    % nb of samples between 2 contiguous buffers  
for ci=1:fix((m-buffer)/shift +1)
    start_index = 1+(ci-1)*shift;
    stop_index = min(start_index+ buffer-1,m);
    time_index(ci) = round((start_index+stop_index)/2);  % time index expressed as sample unit (dt = 1 in this simulation)
    rms_data(ci) = my_rms(data(start_index:stop_index));  % 

end
x_rms = x(time_index);
figure(1),
plot(x,data,x_rms,rms_data,'r*');

%select index if they belong to ranges : 0 - 33% of max / 33 - 66% of max / above 66% of max
ind1 = find(rms_data<=max(rms_data)/3); % range : 0 - 33% of max
ind2 = find(rms_data>max(rms_data)/3 & rms_data<=max(rms_data)*2/3); % range : 33 - 66% of max
ind3 = find(rms_data>max(rms_data)*2/3); % range : above 66% of max
figure(1),
plot(x,data,'k',x_rms(ind1),rms_data(ind1),'g*',x_rms(ind2),rms_data(ind2),'b*',x_rms(ind3),rms_data(ind3),'r*');



%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
function x_rms = my_rms(x)

    x_rms = sqrt(mean(x.^2));
    
end

Answer 2

[Mathieu NOE]

hello Nina

I was about to reply on your other post when it disappeared : did you delete it ?

For your info, this was the demo code I have been preparing for you , maybe still of interest :

clearvars
% dummy data
n=1000;
x=linspace(0,2*pi*3,n);
data = 0.25*ones(size(x));
data = max(data,-cos(x).*(1+x/10));
data(300:303) = 2; % add a spike (for fun)

%% parameters
min_contiguous_samples = 10; % consider "red" segments only if they are at least this length (and contiguous)

% running rms (buffered) parameters : 
buffer = 100;  % nb of samples in one buffer (buffer size) 
overlap = buffer-1; % overlap expressed in samples

%% main loop %%%%
    m = length(data);
    shift = buffer-overlap;    % nb of samples between 2 contiguous buffers  
    for ci=1:fix((m-buffer)/shift +1)
        start_index = 1+(ci-1)*shift;
        stop_index = min(start_index+ buffer-1,m);
        time_index(ci) = round((start_index+stop_index)/2);  % time index expressed as sample unit (dt = 1 in this simulation)
        rms_data(ci) = my_rms(data(start_index:stop_index));  % 

    end
    x_rms = x(time_index);

    %select index if they belong to ranges : 0 - 33% of max / 33 - 66% of max / above 66% of max
    ind1 = (rms_data<=max(rms_data)/2); % range : 0 - 50% of max
%     ind2 = find(rms_data>max(rms_data)/2 & rms_data<=max(rms_data)*2/3); % range : 33 - 66% of max
    ind2 = (rms_data>max(rms_data)/2); % range : 50 - 100% of max (why limit the upper value ??)
    
    % now define start en end point of "red" segments
    [begin2,ends2] = find_start_end_group(ind2);
    length_ind2 = ends2 - begin2;
    ind22= length_ind2>min_contiguous_samples; % check if their length is valid (above min_contiguous_samples value)
    begin2 = begin2(ind22); % selected  points 
    ends2 = ends2(ind22); % selected  points 
    
    % define for plot the red / green rms data 
    x1 = x_rms(ind1);
    rms_data1 = rms_data(ind1);
    x2 = x_rms(ind2);
    rms_data2 = rms_data(ind2);
    % define the begin / ending x, y values of raw data 
    x2_begin = x_rms(begin2);
    data_begin = interp1(x,data,x2_begin);
    x2_ends = x_rms(ends2);
    data_ends = interp1(x,data,x2_ends);
    
figure(1),
plot(x,data,'k',x1,rms_data1,'g.',x2,rms_data2,'r.',x2_begin,data_begin,'dc',x2_ends,data_ends,'dm','MarkerSize',12);
legend('signal','rms below 50%','rms above 50%','begin points','end points');

% store each "red" segment separately (in cell array)
figure(2), hold on
for ci = 1:length(begin2)
    ind = (x>=x2_begin(ci) & x<=x2_ends(ci))
    xx = x(ind);
    yy = data(ind);
    data_store{ci} = [xx(:) yy(:)]; % 2 columns : time / data
    plot(xx,yy);
end
hold off

%%%% end of main file %%%%%%%%%%%

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
function [begin,ends] = find_start_end_group(ind)
% This locates the beginning /ending points of data groups
    D = diff([0,ind,0]);
    begin = find(D == 1);
    ends = find(D == -1) - 1;
end

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
function x_rms = my_rms(x)

    x_rms = sqrt(mean(x.^2));
    
end

Answer 3

[Image Analyst]

In case you delete again, here is the current version of the question:

---------------------------------------------------------------------------------------------------

Hi,

I have data. I have segmented the signal in windows and calculated rms. Can you please help in finding multiple minimum values for this signal?

[v, i] = min(rms);

Thank you!

---------------------------------------------------------------------------------------------------

So to do that, you can do

[sortedRMS, indexes] = sort(rms, 'ascend');

the indexes will tell you the indexes for the RMS with the smallest rms value first, at indexes(1), and the largest rms value last, at indexes(end).