Getframe function problem: Captured Image will not re-display at same size

goal: I want to be able to capture the color by pixel location data (C-data?) of any current figure, scramble the locations of any black pixels in it, and then re-display the figure in scrambled form.

problem: After scrambling the figure and re-displaying it, it is now a different size than the old figure. I want it to be the same size as the unscrambled version.

I think the problem lies in how I am capturing and then displaying the image. This is how I am doing that: (after creating a current figure). Attached is a screenshot of the problem for clarity. I am running 2015a.

function[CurrentFigureData, CurrentFigureMap]= WriteCF2Workspace()

F = getframe(gcf);

[CurrentFigureData, CurrentFigureMap] = frame2im(F);

end

figure(2);ScreenDims = get(0, 'ScreenSize'); set(2,'position',ScreenDims, 'menubar','none'); image(CurrentFigureData); axis off;

[Image Analyst]

Why not just use an axes control with an image in it, instead of trying to do the whole figure?

[Michael Bronstein]

Thanks for the idea. Could you explain how making that change would solve the problem? -Michael

Answer 1

[Walter Roberson]

curfig = gcf;
curfig_units = get(curfig, 'units');
curfig_size = get(curfig, 'Position');

curimage_h = findall(curfig, 'type', image');
curax = ancestor(curimage_h, 'axes');

fig2 = figure('Units', curfig_units, 'Position', curfig_size);
ax2 = copyobj(curfig_ax, fig2);
image2 = copyobj(curimage_h, ax2);

[Michael Bronstein]

Thanks for the help so far Walter! This looks promising, but I have one question. When I run that code I get the following error: Error using copyobj Copyobj cannot create a copy of an invalid handle.

Error in Untitled (line 7) ax2 = copyobj(curfig_ax, fig2);

Do you know why that might be?

[Walter Roberson]

curfig = gcf;
curfig_units = get(curfig, 'units');
curfig_size = get(curfig, 'Position');

curimage_h = findall(curfig, 'type', image');
curax = ancestor(curimage_h, 'axes');

fig2 = figure('Units', curfig_units, 'Position', curfig_size);
ax2 = copyobj(curax, fig2);
image2 = copyobj(curimage_h, ax2);

[Michael Bronstein]

Unfortunately, this solution does not work. When I insert the scrambled image on the figure that this script creates it does not align with the original.

[yuval]

Hi Michael, I'm facing similar difficulty, were you able to figure it out?

[Michael Bronstein]

Here is the code I was working on at the time:

function[TransformMe,CurrentFigureData,CurrentFigureMap] = MaskFromCF() %iterate over each dimension from 1:dimsize. Find all indicies where all 3 %= 0. Do it so that you only check dim 2 if dim 1 was 0, and only dim 3 if %dim 2 was 0. Record those indicies, and you will have black pixels! %note that white is 255, 255,255 RGB

%create workspace var with image data [CurrentFigureData, CurrentFigureMap] = WriteCF2Workspace;

%now modify it to make mask, start by finding all black pixels (works only for black text on white) SizeOfData = size(CurrentFigureData); display(SizeOfData(1)) LocationsOfBlack = []; for i = 1:SizeOfData(1) for j = 1:SizeOfData(2) if CurrentFigureData(i,j,1) == 0 && CurrentFigureData(i,j,2) == 0 && CurrentFigureData(i,j,3) == 0 LocationsOfBlack{1,end+1} = {i,j}; end end end %now that you found all black pixels, find the furthest left/right values %that are black. You will scramble the pixels between these values.

%%%%you can mod these parameters to make it look only in a certain %%%%space.%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% furthestX = -inf; smallestX = inf; furthestY = -inf; smallestY = inf;

%the parts after the && signs in this for loop specify screen locations to search in. Remove to search whole screen. %NOTE THAT X AND Y HERE CORRESPOND TO THE OPPOSITE SCREEN Dimension (restricting X effects Y range when shown etc.) %Also note that low numbers correspond to the top of the screen and higher %ones to the bottom of the screen for i = 1:length(LocationsOfBlack) testx = cell2mat(LocationsOfBlack{i}(1)); testy = cell2mat(LocationsOfBlack{i}(2)); if testy > 270 && testy < 1022 && testx > 80 && testx < 270 %restricts screen scan area for mask (by limiting scan for black). Change numbers to alter screen scan area. Go = 1; else Go = 0; end if Go == 1 if cell2mat(LocationsOfBlack{i}(1)) > furthestX furthestX = cell2mat(LocationsOfBlack{i}(1)); end if cell2mat(LocationsOfBlack{i}(1)) < smallestX smallestX = cell2mat(LocationsOfBlack{i}(1)); end if cell2mat(LocationsOfBlack{i}(2)) > furthestY furthestY = cell2mat(LocationsOfBlack{i}(2)); end if cell2mat(LocationsOfBlack{i}(2)) < smallestY smallestY = cell2mat(LocationsOfBlack{i}(2)); end end end %end of section where you can mod parameters to impact search field %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Xdiff = furthestX-smallestX; Ydiff = furthestY-smallestY;

%Move around the pixels ShuffleX = randperm(Xdiff+1)+smallestX; %start shuffling from smallest to largest X. ShuffleY = randperm(Ydiff+1)+smallestY; %you now have the indexes you need to shuffle, now shuffle them in %CurrentFigureData TransformMe = CurrentFigureData; for i = smallestX:furthestX for j = smallestY:furthestY k = i-(smallestX-1); l = j-(smallestY-1); TransformMe(i,j,:) = CurrentFigureData(ShuffleX(k),ShuffleY(l),:); end end

[Michael Bronstein]

And the other function:

function[CurrentFigureData, CurrentFigureMap]= WriteCF2Workspace() F = getframe(gca); [CurrentFigureData, CurrentFigureMap] = frame2im(F); end

[Michael Bronstein]

I think it did end up working, but unfortunately I can't test it atm because I don't have access to my normal pc. Hope the code helps. Good luck!

Answer 2

[yuval]

Replace gcf with gca, i.e.,

F = getframe(gca); [CurrentFigureData, CurrentFigureMap] = frame2im(F);