将MATLAB矩阵另存为特定大小和分辨率的图像

Question

After having read this question, I wanted to use this code to save my picture as to particular size.

    I_c(:,:) = cropped_matrix(i,:,:);
    set(I_c, 'PaperUnits', 'inches');
    x_width = 7.25;
    y_width = 9.125;
    set(I_c, 'PaperPosition', [0 0 x_width y_width]);
    imshow(I_c);
    saveas(I_c,'fig1.pdf');

I_c represents a 2D matrix (about 40x40) of uint8 's.

However, I get the error:

Error using set Invalid handle

This makes me believe that I can only use this code with figures and not matrices which contain matrices. How would I go about this?

I have looked at the API for print , as suggested as the first answer of the aforementioned linked question, but it also suggests using set and 'PaperUnits' .

Note: This question also looks at this problem but suggests the same solution.

---

Notes About Crowley's Answer

1. So I just tried the code that you have given with your answer. The .jpg being produced is the one shown below. How would I get rid of all the extra white area through code?
2. How would I change the colourmap of the figure being produced, to greyscale when I just use image(ImData) ?

And here is how the actual figure appears:

Here is the code that I have put in:

 im = image(I_c); 
 set(gcf,'units','inches','position',[1 2 5 5]);
 set(gca,'ydir','normal','units','centimeters','position',[0 0 0.5 0.5].*get(gcf,'position')) ;
 filename = strcat('slice',int2str(i),'_','bead',int2str(j),'.jpg');
 saveas(im,filename);

Answer 1

Suppose we have matrix I_c containing values and x and y the coordinates so value I_c(ii,jj) corresponds to x(ii) and y(jj) coordinates.

Then:

ImMin=min(min(I_c));    % Find the minimum value in I_c
ImData=I_c-ImMin;       % Ensure non-zero values
ImMax=max(max(ImData)); % Find maximum value in ImData
ImData=ImData./ImMax;   % Ensure ImData do NOT exceed 1

image(x,y,ImData*[1 1 1]) % plot the image in greyscale
set(gcf,'units','inches','position',[1 2 5 5]) % set active figure properties
set(gca,'ydir','normal','units','inches','position',[0 0 1 1].*get(gcf,'position')) % set active axes properties

export_fig(gcf,'figure-I_c','-png','-nocrop')

'ydir','normal' parameter change default (1,1) point being in top-left corner to "normal" position in bottom-left corner.
[0 0 1 1].*get(gcf,'position) will read active figure position (here [1 2 5 5]) and after element-by-element multiplication the [0 0 5 5] is passed to the position which causes that axes fit the image.

export_fig function will create figure-I_c.png image as it is shown in Matlab figure, if -nocrop is omitted the possible white space in the edges is cropped out. This function is available from MathWorks' File Exchange .