Steven's answer is probably the core of what you need, but I'm going to throw this out there.
This may or may not be relevant, but given that you're describing the data as an "image", bear in mind that representing the data with contour() will display it with a different YDir than might be expected of an image as displayed using image(), imagesc(), or imshow().
inpict = imread('cameraman.tif'); % a visually recognizable image
inpict = imgaussfilt(inpict,5); % make it smooth for a cleaner demo
inpict = im2double(inpict)*10 + 100; % put it on some arbitrary scale
imagesc(inpict) % origin is in the NW corner
figure
contour(inpict) % origin is in the SW corner
If you find that you need to also reorient the visualization so that they match, you can use some combination of setting the 'YDir' property of the parent axes, and using flip(..., 1) or flipud() on the data.
