Image Processing Reference
In-Depth Information
2.1.2 Image Segmentation
Segmentation is one of the most important steps in medical image processing
that is done normally after enhancement. It extracts any clot/abnormal lesion
or blood cells/blood vessels present in an image. Each region in a segmented
image possesses homogeneous properties with respect to features, such as
grey level, texture or colour, and the property is different for different regions
in an image. It partitions the image into disjoint sets corresponding to objects
in the scene, which are very important in identifying different types of leu-
kocytes or counting blood vessels or finding the size of the tumour or any
other abnormalities present in the human body. After segmentation, analysis
is carried out and is explained in detail in a separate chapter.
2.1.3 Boundary Detection
Boundary detection is an important process in medical image processing. It
finds the structural information of the image, thus drastically reducing the
data to be processed. In an image, there are changes in intensity levels. Edge
is present when there is a change in grey level intensity. Edge image may be
termed as gradient image. The boundaries indicate the location and the shape
of the objects such as abnormal lesions or tumour or blood cells/vessels or
any other structures present in an image. The edges in medical images are
not distinct due to poor contrast and unequal illumination. So initially, the
edges are sometimes enhanced before performing edge detection, which is
explained in detail in a separate chapter.
2.1.4 Morphology
Morphology is a non-linear image processing technique that deals with the
shape of the image features. Sometimes, there is a need to obtain a skeleton of
an image or to dilate/erode an image or close/open an image to remove or fill
any holes or gaps present in the image. It may be used to find the gradient of an
image. It uses a structuring element of any size and shape which is applied on
an image to perform operation. It is explained in detail in a separate chapter.
2.2 Image Registration
Image registration is the process of finding the transformation that aligns or
maps an object of one coordinate to an object of a different coordinate. The
goal of registration is to find a corresponding anatomical or functional posi-
tion in two or more images. Maurer and Fitzpatrick [13] defined image regis-
tration as 'It is the one-to-one mapping between the coordinates in one space
and those in another, such that the points in two spaces that correspond
Search WWH ::
Custom Search