Dilation - Erosion - Very Good PDF

21/11/2016 Binary Image Operations (Image Processing Toolbox)
Dilation and Erosion
The main morphological operations are dilation and erosion. Dilation and erosion are related operations, although they produce very different results. Dilation
adds pixels to the boundaries of objects (i.e., changes them from off to on), while erosion removes pixels on object boundaries (changes them from on to off).
Each dilation or erosion operation uses a specified neighborhood. The state of any given pixel in the output image is determined by applying a rule to the
neighborhood of the corresponding pixel in the input image. The rule used defines the operation as a dilation or an erosion.
For dilation, if any pixel in the input pixel's neighborhood is on, the output pixel is on. Otherwise, the output pixel is off.
For erosion, if every pixel in the input pixel's neighborhood is on, the output pixel is on. Otherwise, the output pixel is off.
The neighborhood for a dilation or erosion operation can be of arbitrary shape and size. The neighborhood is represented by a structuring element, which is a
matrix consisting of only 0's and 1's. The center pixel in the structuring element represents the pixel of interest, while the elements in the matrix that are on (i.e.,
= 1) define the neighborhood.
The center pixel is defined as floor((size(SE)+1)/2), where SE is the structuring element. For example, in a 4by7 structuring element, the center pixel is (2,4).
When you construct the structuring element, you should make sure that the pixel of interest is actually the center pixel. You can do this by adding rows or
columns of 0's, if necessary. For example, suppose you want the neighborhood to consist of a 3by3 block of pixels, with the pixel of interest in the upperleft
corner of the block. The structuring element would not be ones(3), because this matrix has the wrong center pixel. Rather, you could use this matrix as the
structuring element.
0 0 0 0
0 1 1 1
0 1 1 1
0 1 1 1
For erosion, the neighborhood consists of the on pixels in the structuring element. For dilation, the neighborhood consists of the on pixels in the structuring
element rotated 180 degrees. (The center pixel is still selected before the rotation.)
Suppose you want to perform an erosion operation. Figure 91 shows a sample neighborhood you might use. Each neighborhood pixel is indicated by an x, and
the center pixel is the one with a circle.
http://radio.feld.cvut.cz/matlab/toolbox/images/binary3.html 1/3
Figure 91: A Neighborhood That Will Represented as a Structuring Element
The structuring element is therefore
1 0 0 0 0
0 1 0 0 0
0 0 1 0 0
0 0 0 1 0
0 0 0 0 1
The state (i.e., on or off) of any given pixel in the output image is determined by applying the erosion rule to the neighborhood pixels for the corresponding
pixel in the input image. For example, to determine the state of the pixel (4,6) in the output image:
Overlay the structuring element on the input image, with the center pixel of the structuring element covering the pixel (4,6).
Look at the pixels in the neighborhood of the input pixel. These are the five pixels covered by 1's in the structuring element. In this case the pixels are:
(2,4), (3,5), (4,6), (5,7), (6,8). If all of these pixels are on, then set the pixel in the output image (4,6) to on. If any of these pixels is off, then set the pixel
(4,6) in the output image to off.
You perform this procedure for each pixel in the input image to determine the state of each corresponding pixel in the output image.
Note that for pixels on borders of the image, some of the 1's in the structuring element are actually outside the image. These elements are assumed to
cover off pixels. (See the earlier section, Padding of Borders.) As a result, the output image will usually have a black border, as in the example below.
The Image Processing Toolbox performs dilation through the dilate function, and erosion through the erodefunction. Each of these functions takes an input
image and a structuring element as input, and returns an output image.
This example illustrates the erosion operation described above.
BW1 = imread('circbw.tif');
SE = eye(5);
BW2 = erode(BW1,SE);
imshow(BW1)
figure, imshow(BW2)
Figure 92: Circbw.tif Before and After Erosion with a Diagonal Structuring Element
Notice the diagonal streaks in the output image (on the right). These are due to the shape of the structuring element.

Dilation - Erosion - Very Good PDF

Uploaded by

Copyright:

Available Formats

Dilation - Erosion - Very Good PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Dilation - Erosion - Very Good PDF

Uploaded by

Copyright:

Available Formats

21/11/2016 Binary Image Operations (Image Processing Toolbox)

You might also like