{"id":365,"date":"2011-03-22T22:34:12","date_gmt":"2011-03-23T02:34:12","guid":{"rendered":"https:\/\/blogs.mathworks.com\/steve\/2011\/03\/22\/too-much-information-about-the-size-function\/"},"modified":"2019-10-29T16:32:31","modified_gmt":"2019-10-29T20:32:31","slug":"too-much-information-about-the-size-function","status":"publish","type":"post","link":"https:\/\/blogs.mathworks.com\/steve\/2011\/03\/22\/too-much-information-about-the-size-function\/","title":{"rendered":"Too much information about the size function"},"content":{"rendered":"
\r\n

Today's post is proof that this old MATLAB dog can learn new tricks. The subject is this: Given an image array, how can you\r\n most easily determine the number of rows and columns it has? (In code, that is; I don't mean by looking at the Workspace Browser!)\r\n <\/p>\r\n

I'm going to take an indirect path to get there:<\/p>\r\n

1. I'll explain a general principle of MATLAB function behavior.<\/p>\r\n

2. I'll show the that the size<\/tt> function violates this principle.\r\n <\/p>\r\n

3. I'll show how a common use of size<\/tt> can lead you astray.\r\n <\/p>\r\n

4. I'll show the code I used to use.<\/p>\r\n

5. Finally I'll show how to do it a new \"trick\" (that Loren<\/a> taught me) which uses a fairly recent addition to the MATLAB language.\r\n <\/p>\r\n

There's nothing profound here, but there are some subtle issues about using size<\/tt> with multidimensional arrays that image processing users should probably be aware of. (I'll warn you here --- if you're already\r\n bored with this topic, the rest of the post doesn't get better. You should probably just go read one of the other MATLAB Central blogs<\/a> instead.)\r\n <\/p>\r\n

The general principle of MATLAB function behavior is about functions with multiple output arguments. For example, the Image\r\n Processing Toolbox function bwlabel<\/tt> has two output arguments. A sample call using both output arguments is:\r\n <\/p>

  [L,n] = bwlabel(BW);<\/pre>
The first output argument is the label matrix, and the second argument is the number of labeled objects.<\/p>\r\n   
For most MATLAB functions, the meaning of each output argument does not depend on how many output arguments are specified\r\n      when the function is called. In the example above, if you call bwlabel<\/tt> with only one output argument, that output is still the label matrix.  The second output argument is silently dropped by\r\n      MATLAB. And if you call the function with no output argument, the first output is still the label matrix and is used to set\r\n      the automatic workspace variable ans<\/tt>.\r\n   <\/p>\r\n   
I can think of three kinds of exceptions to this behavior. In the first case, common to many graphics functions, a function\r\n      returns something (like a handle) if called with an output argument, but it returns nothing if called without an output argument.\r\n      Examples in this category include plot<\/tt> and bar<\/tt>.\r\n   <\/p>\r\n   
The second kind of exception is for functions that \"pretty-print\" something to the Command Window if called without an output\r\n      argument and return a value if called with an output argument. Examples include dir<\/tt> and whos<\/tt>.\r\n   <\/p>\r\n   
The third category is (unfortunately) the \"everything else\" exceptions. And size<\/tt> is in this group.\r\n   <\/p>\r\n   
You can see the exceptional behavior with your plain ol' two-dimensional matrices. Consider the following two calls to size:<\/p>
 A = rand(3,4);  % A is 3-by-4\r\n m = size(A);\r\n [m,n] = size(A);<\/pre>
m<\/tt> is not the same in both cases. In the first case, size<\/tt> returns the size vector, [3 4]<\/tt>. In the second case, m<\/tt> is the number of rows (3) and n<\/tt> is the number of columns (4).\r\n   <\/p>\r\n   
For higher dimensional arrays it gets more interesting. The rule is this: If you call size<\/tt> with two or more output arguments, the last output argument is the number of elements in that array dimension and all the remaining dimensions of the array<\/i>.\r\n   <\/p>\r\n   
It's probably easier to just show you. Let's look at a 3-by-4-by-5 array and calls to size<\/tt> with a different number of output arguments.\r\n   <\/p>
B = rand(3,4,5);\r\n[m,n,p] = size(B)<\/pre>
\r\nm =\r\n\r\n     3\r\n\r\n\r\nn =\r\n\r\n     4\r\n\r\n\r\np =\r\n\r\n     5\r\n\r\n<\/pre>
Not surprising. Now just call it with just two output arguments:<\/p>
[m,n] = size(B)<\/pre>
\r\nm =\r\n\r\n     3\r\n\r\n\r\nn =\r\n\r\n    20\r\n\r\n<\/pre>
Now the second output is 20 instead of 4! That's because 4*5 is 20.  For fun, call size with 4 outputs:<\/p>
[m,n,p,q] = size(B)<\/pre>
\r\nm =\r\n\r\n     3\r\n\r\n\r\nn =\r\n\r\n     4\r\n\r\n\r\np =\r\n\r\n     5\r\n\r\n\r\nq =\r\n\r\n     1\r\n\r\n<\/pre>
MATLAB is perfectly happy to treat B<\/tt> as a four-dimensional array whose last dimension has size 1. That's actually pretty useful, even essential, for writing reasonable-looking\r\n      multidimensional code in MATLAB.\r\n   <\/p>\r\n   
Now let me connect this back to the problem I posed at the very beginning: find the number of rows and columns in an image\r\n      array. Let's make it a concrete example:\r\n   <\/p>
neb = imread('ngc6543a.jpg'<\/span>);\r\nimshow(neb, 'InitialMagnification'<\/span>, 50)\r\ntitle('Obligatory image'<\/span>)\r\nxlabel('This blog is supposed to be about image processing after all'<\/span>)<\/pre> 
Here's the naive attempt to ask for the number of rows and columns in our image:<\/p>
[nrows,ncols] = size(neb)<\/pre>
\r\nnrows =\r\n\r\n   650\r\n\r\n\r\nncols =\r\n\r\n        1800\r\n\r\n<\/pre>
But unfortunately that's not right because neb<\/tt> is a 3-D array. Therefore the second output is the number of columns times 3, the number of color planes. There are two ways\r\n      I've always used to make sure I get the right answer. The first is to call size<\/tt> with three output arguments, even though I only need two:\r\n   <\/p>
[nrows,ncols,junk] = size(neb)<\/pre>
\r\nnrows =\r\n\r\n   650\r\n\r\n\r\nncols =\r\n\r\n   600\r\n\r\n\r\njunk =\r\n\r\n     3\r\n\r\n<\/pre>
I always found this vaguely unsatisfactory because I don't like unused variables cluttering up my code. The second way requires\r\n      two separate calls to size<\/tt>:\r\n   <\/p>
nrows = size(neb,1)<\/pre>
\r\nnrows =\r\n\r\n   650\r\n\r\n<\/pre>
ncols = size(neb,2)<\/pre>
\r\nncols =\r\n\r\n   600\r\n\r\n<\/pre>
Loren<\/a> pointed out to me recently that a feature added to R2009b, using the tilde character in input and output argument lists,\r\n      can be used here. Used in an output argument list, the tilde character says to MATLAB, \"Hey, I'm not going to need this output\r\n      argument and don't want to even give it a variable name, so please just clean it up automatically for me.\" Loren and I both\r\n      posted about this feature (hers<\/a> and mine<\/a>) if you want to go read more about it. As I recall, not everyone was a fan. Oh, well. I like it.\r\n   <\/p>\r\n   
Anyway, here's how you can use it in the call to size<\/tt>:\r\n   <\/p>
[nrows,ncols,~] = size(neb)<\/pre>
\r\nnrows =\r\n\r\n   650\r\n\r\n\r\nncols =\r\n\r\n   600\r\n\r\n<\/pre>
Maybe this was more than you really wanted to know about size<\/tt>. However, I think that if you regularly write code that handles multidimensional arrays, eventually you'll be happy that\r\n      you made it all the way to the end.\r\n   <\/p>