GIF, JPEG, etc., are image FILE format. Image data may be compressed, and additional information is added into the file, e.g., BITMAP header data. But once an image is loaded into the MATLAB workspace, it's just a normal matrix or array: for a gray-scale image, a matrix, and for an RGB image, a 3D array, with the size of the third dimension being 3.
There are several aspects indicating that a matrix or a 3D array be more likely to be image data. If the data type is uint8, or the data type is double and the values of the elements range between 0 and 1, the data are more likely to be an image commonly encountered during image processing.
But basically speaking, any matrix or 3D array (with the size of the 3rd dimension being 3) can be "treated" or "interpreted" or "shown" in the form of an image. This is a very common processing of the data for visualization. For example, given A as a matrix corresponding to the values of a function f(x,y) over a rectangular grid, you can use
imshow(A,[])
to show the data as a gray-scale image, or even better, use
imshow(A,[]), colormap(a_colormap)
to show the data as a pseudo-color image. You can even use imwrite to save the data as an image file. But all these do not necessarily mean that the matrix A itself is originally image data.
