Borrar filtros
Borrar filtros

How to run the following function on GPU or make it Faster

3 visualizaciones (últimos 30 días)
Med Future
Med Future el 19 de Jun. de 2023
Comentada: Med Future el 19 de Jun. de 2023
I have following MATLAB function It is very slow it takes so much time,it takes 5 seconds to run, but i want to run it in miliseconds,
Can anyone Help me running this code on GPU. I have also Attached the Dataset Below
[valueestimationimage ] = Parameterestimate(Batchdata)
fig=figure; set(fig,'visible','off');
h=histogram(Batchdata,10000,"BinMethod","sturges",'BinWidth',1,'BinLimits',[1 10000]);
sumofbins=max(h.Values);
% size_MP=round(10/100*sumofbins);
size_MP=round(10/100*sumofbins);
ValueofHistogram= h.Values;
Bindata=h.Data;
Binedges=h.BinEdges;
Binedges(end) = Inf;
deleted_data_idx = false(size(Bindata));
for i=1: length(ValueofHistogram)
if ValueofHistogram(i)<size_MP;
deleted_data_idx(Bindata >= Binedges(i) & Bindata < Binedges(i+1)) = true;
end
end
close(fig);
Bindata(deleted_data_idx) = [];
fig=figure; set(fig,'visible','off');
Freq_Data = Bindata;
h = histogram(Freq_Data, 10000, "BinMethod", "sturges", 'BinWidth', 1, 'BinLimits', [1 10000]);
[N, Edges, Bin] = histcounts(Freq_Data, 10000, "BinMethod", "sturges", 'BinWidth', 1, 'BinLimits', [1 10000]);
Retain = N > max(N) / 3.5;
% Find the bin indices that satisfy the condition
FindBins = find(Retain);
% Update RetainDataLv based on the valid bin indices
RetainDataLv = ismember(Bin, FindBins);
% Apply the logical indexing to retrieve the corresponding data
Bindata = Freq_Data(RetainDataLv);
close(fig);
Bindata=round(Bindata).';
[GC, GR] = groupcounts(Bindata) ;
countThresh =30 ; % change this untill you see that the data is fully denoised
denoisedData = Bindata(ismember(Bindata, GR(GC>countThresh))) ;
% incomingdata= denoisedData.';
if isempty(denoisedData)
incomingdata=Bindata.';
else
incomingdata=denoisedData.';
end
[row, column] = size(incomingdata);
for eachrow=1:row
if column>=1000
% buffered(eachrow,:) = buffer(incomingdata, 1000);
groupsize = 1000;
sig = incomingdata(:);
if isempty(sig)
error('signal is empty, cannot buffer it');
end
sigsize = numel(sig);
fullcopies = floor(groupsize ./ sigsize);
sig = repmat(sig, 1+fullcopies, 1);
sigsize = numel(sig);
leftover = mod(sigsize, groupsize);
if leftover ~= 0
sig = [sig; sig(1:groupsize-leftover)];
end
buffered = buffer(sig, groupsize);
else
targetsize = 1000;
sizeofincomingdata = column;
nrep = targetsize / sizeofincomingdata;
fullrep = floor(nrep);
leftover = targetsize - fullrep * sizeofincomingdata;
buffered=[repmat(incomingdata(eachrow,:), 1, fullrep), incomingdata(1:leftover)];
sig=buffered.';
end
end
signal=sig.';
[numImages, lenImage] = size( signal);
imbg = false(10000,lenImage); % background "color"
imfg = ~imbg(1,1); % forground "color"
imSizeOut=[10000 lenImage];
% ImageSize
for k= 1:numImages
imData = round( signal(k,:)); % get pattern
[~,Y] = meshgrid(1:lenImage,1:10000); % make a grid
% black and white image
BW = imbg;
BW(Y==imData)=imfg;
valueestimation=imbinarize(imresize(uint8(BW),imSizeOut));
% convert to uint8 (0 255)
valueestimationimage = im2uint8(valueestimation);
% resize (from 1000x1000)
SE=strel('disk',2);
BW=imdilate(BW,SE);
BW=imbinarize(imresize(uint8(BW),imSizeOut));
% convert to uint8 (0 255)
imoriginalestimate = im2uint8(BW);
end
end

Respuestas (1)

Lakshya
Lakshya el 19 de Jun. de 2023

Categorías

Más información sobre GPU Computing en Help Center y File Exchange.

Productos


Versión

R2022a

Community Treasure Hunt

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

Start Hunting!

Translated by