08-16-2017, 10:22 PM
introduction
Image segmentation can be defined as the process of
dividing the image on non-intersecting Uniform
regions. These homogeneous areas, as a rule, we Have
Similar items of interest or A part of them.
homogeneity of segmented areas can be
Measurements shall be carried out by using some image properties (e. g. Pixel
The Intensity). On the other hand, clustering can be
is defined as the optimal partitioning of the specified set p
the data points in the subgroups so that data points
Belonging to the same group as similar to each
While other data points as possible in two
different groups share the maximum difference.
Segmentation of an image can be seen as clustering
Where the problem is Describing each pixel
match the pattern match, and each region of the image (i.e.,
Segment) corresponds to the cluster. Therefore, many
Clustering algorithms are widely used for decision
the segmentation problem
Fuzzy c-means (FCM) seems to be the most
popular algorithm for fuzzy clustering.
Many researchers have attempted to Modify
Classic TCM and application to image segmentation
in the past few years [7-12].
However, most of them upgrade an existing cluster
Algorithms assume a priori knowledge on the number
classes with, and in many practical situaciones,
The number of classes is unknown or
It is impossible to determine even approximate.
Searching for the optimal number of clusters in a large
The DataSet is a challenging task, we generally. several
The researchers studied the problem.
However, the result is still not satisfactory.
Automatic clustering of evolutionary
Strategy (ES), evolutionary programming (EP) and
The variable-length string is the genetic algorithm (VGA) Have
Reported in.
Compared with the huge number of clustering algorithms
based on various types of evolutionary algorithms
not a lot of work, it was reported on algorithms, as
Particle Swarm Optimization (PSO) or differential
Evolution (DE) to cluster pixels. to
As far as we know, the PSO is used for average user
Automatic fuzzy clustering to date.
In this paper, we used a modified version of the
The basic PSO algorithm for fuzzy segmentation
images. Because the pixels in an image, usually
high correlation, we incorporated the spatial
Information about each pixel of the marks in
different clusters. We borrowed the concept of spatial
Information from that used it only for TCM
The Algorithm. However, the automatic clustering scheme
as well as multi-elitist model of PSO are new. Us
Be Among our comparison method and very
The last automatic fuzzy clustering technique in terms of the accuracy of the clustering, the receive rate
solution and algorithm of reliability