Digital Signal Processing Reference
In-Depth Information
ability, but also the maximum extent possible to retain the characteristics of the
original image [4-5].
In this paper, with studying composite morphological edge detection method, we
propose an improved composite morphological edge extraction method. Using
Composite morphological features to improve the morphological gradient operator,
and synthesize the final edge with the adaptive weights, the simulation results show
that the method has a strong ability to suppress noise, the extracted edge is more
smooth, consistent and complete.
If you have more than one surname, please make sure that the Volume Editor knows
how you are to be listed in the author index.
2
Improved Multi-structure and Multi-faceted Morphology
Edge Detection Method
2.1
Basis for the Work
Method of morphological edge extraction is achieved by using morphological
gradient, it can get the image edge on the basis of retaining the image texture features
information very well. Common morphological gradient operator is defined as:
g
=
(
f
⊕
b
)
−
(
f
Θ
b
)
(1)
As to extract better edge in the image contains more noise, we have introduced the
anti-noise morphological gradient operator, such as anti-noise dilation and erosion
morphological gradient operator, which is defined as:
g
=
(
f
b
)
⊕
b
−
(
f
•
b
)
Θ
b
(2)
4
The choice of structural elements in morphology will directly affect the image
processing results. In the actual image, the edge information is very rich. It is difficult
to get the edge full detected in morphological operations using a single fixed structure
element, so composite morphological methods can be used to extract the image
edges, common-used composite morphology include multiple-scale morphology,
multi-structure morphology and multi-faceted morphology [6].
Different composite morphological methods have specific algorithms, but the basic
idea is basically the same. That is applying different size, shape or orientation of the
structural elements to morphological operations synthetically, finally we synthesis the
final edge by average weights.
The following pictures are the edge extraction results of the three composite
morphological method, the selected image is one of the NASA AVIRIS image.
Equations should be punctuated in the same way as ordinary text but with a small
space before the end punctuation mark.
