Last week I wrote about our desire to reduce the memory used by bwlabel. I also wrote about other issues we hoped to address:
- The label matrix uses memory proportional to the total number of image pixels. This seems unreasonable when you're working with an image containing only a few object pixels.
- When bwlabel was used with regionprops (a frequent combination), some computational work was duplicated in both functions.
- Many users never use the label matrix directly. They only compute it in order to pass it to regionprops. That implies there may be a simpler, more usable interface design.
With so much code out there using bwlabel and regionprops, though, we did not want to introduce any compatibility problems.
Our solution was to introduce two new functions in R2009a, bwconncomp and labelmatrix, as well as a couple of new syntaxes for the existing function regionprops.
The new function bwconncomp finds the connected components of a binary image, just as bwlabel does. But it does not return the result as a label matrix.
BW = [1 1 1 0 0 0 0 0
1 1 1 0 1 1 0 0
1 1 1 0 1 1 0 0
0 0 0 0 0 0 1 0
0 0 0 0 0 0 1 0
0 0 0 0 0 0 1 0
0 1 1 0 0 1 1 0
0 1 1 0 0 0 0 0];
s = bwconncomp(BW)
s =
Connectivity: 8
ImageSize: [8 8]
NumObjects: 3
PixelIdxList: {[9x1 double] [4x1 double] [9x1 double]}
As you can see, bwconncomp produces a structure. The critical information for subsequent processing is the PixelIdxList field, which contains the linear indices of the pixels belonging to each object. For example, the 2nd object found comprises four pixels with the following linear indices:
s.PixelIdxList{2}
ans =
15
16
23
24
A key aspect of this design is that the amount of memory required is proportional to the number of object pixels, not the number of image pixels. For a 1000-by-1000 image containing a single foreground pixel, the output data structure is nicely compact:
BW2 = false(1000, 1000); BW2(500, 500) = true; s2 = bwconncomp(BW2)
s2 =
Connectivity: 8
ImageSize: [1000 1000]
NumObjects: 1
PixelIdxList: {[499500]}
whos s2Name Size Bytes Class Attributes s2 1x1 596 struct
You can compute the label matrix, if you really need it, using the new function labelmatrix.
L2 = labelmatrix(s2);
whos L2Name Size Bytes Class Attributes L2 1000x1000 1000000 uint8
Notice that L2 is a uint8 matrix. Unlike bwlabel, labelmatrix does adjust the output data type based on the number of objects.
The last related change was to regionprops. As I mentioned before, we've noticed that this is a very common coding pattern:
L = bwlabel(bw); s = regionprops(L, ...);
We've modified the syntax for regionprops so that you can now do this in a one step:
s = regionprops(bw, ...);
For example:
bw = imread('text.png'); s = regionprops(bw, 'Centroid');
These several new and modified functions, working together, nicely solve several design problems:
- Reducing memory use
- Increasing speed (in many cases) by eliminating duplicate computation
- Converting a two-step workflow to a one-step workflow
We had one last concern, though. It takes code changes to benefit from these enhancements. Users who aren't aware of the new features and don't change their code will never benefit. So what could we do to help users discover the new features?
I'll comment about that next time.
Get
the MATLAB code
Published with MATLAB® 7.8
Great!
I had to encode (and later decode) my labes matrix as a sparse matrix due its large size and conequent impact on run-time memory!
Great idea!
I previously had to use sparse matrices in order to save the labels matrix(when passing the labels matrix on the way) because of the huge run-time memory it consumed.
The update for ‘regionprops’ to calculate the perimeter of each labeled region in a label matrix is very useful. But is there any property of ‘regionprops’ to get the boundary arrays (Example: I need the output of B in the source code given by
B = regionboundariesmex(double(L),8);)
SlamJ—You could get the ‘Image’ property using regionprops and then call bwboundaries on each.
Thanks for ur quick reply. ‘Image’ property using regionprops and calling bwboundaries on each gives the boundary on each region separately . But I need boundaries of each labelled region on a whole.(Example: label matrix with 420 objects should give the boundaries of all 420 objects as a whole and not taking each object separately to give each separated boundary)
error: Function BWBOUNDARIES expected at most 4 input arguments but was called instead with 420 input arguments.
Can u please explain the basic idea behind the regionboundariesmex.cpp like how does it search across the image for the first dissimilar pixel and use that location of the pixel as the boundary pixel and how does it trace the remaining?
SlamJ—Sounds to me like the convex hull might be of interest to you. See convhull. I think you may be trying to use regionprops and bwboundaries for something they are not really intended for.
Jestine—I might do a post in the future on contour tracing. In the meantime, take a look at tutorial number 014, “Contour tracing,” on the imageprocessingplace.com tutorials page.