30-05-2012, 11:12 AM
AUTOMATED LOCALISATION OF RETINAL OPTIC DISK USING HOUGH TRANSFORM
[attachment=23456]
INTRODUCTION
Fundus images are used for diagnosis by trained clinicians
to check for any abnormalities or any change in the retina.
They are captured by using special devices called
ophthalmoscopes. A typical fundus image with its features
marked is shown in the Figure 1. Each pixel in the fundus
image consists of three values namely red, green and blue,
each value being quantised to 256 levels.
The information about the optic disk can be used to
examine severity of some diseases such as glaucoma.
Changes in the optic disk can indicate the current state and
progression of a certain disease. The location of the optic
disk is an important issue in retinal image analysis as it is a
significant landmark feature, and its diameter is usually used
as a reference length for measuring distances and sizes.
METHODOLOGY
The optic disk appears in colour fundus images as a bright
yellowish or white region. Its shape is more or less circular,
interrupted by outgoing vessels. Although sometimes due to
the nature of the photographic projection it has the form of
an ellipse.
Mathematical morphology in image processing is
particularly suitable for analysing shapes in images. The two
main processes are those of dilation and erosion. These
processes involve a special mechanism of combining two
sets of pixels. Usually, one set consists of the image being
processed and the other a smaller set of pixels known as a
structuring element or kernel.
RESULTS
The proposed technique was tested on the DRIVE database
of retinal images [8]. This consists of 40 fundus images of
dimensions 768×584, captured by a Canon CR5 nonmydriatic
3CCD camera with a 45o field of view (FOV).
These images contain both normal (healthy) and abnormal
retinas. In this study 38 of these images were used (2 images
have been excluded for not having visually-detectable optic
disks) consisting of 31 normal and 7 abnormal images. The
performance of the optic disk localisation was evaluated
based on the determined optic disk location with regard to
an expert. Our proposed method is capable of localising the
optic disk correctly for 36 of these images. The success rate
was 94.7%. This was a notable result compared to the
method used in [7] achieving 90.25% for the same DRIVE
dataset.
CONCLUSIONS
In this paper, a computer-assisted retinal image analysis
system for the localisation of optic disk in colour digital
fundus images has been presented. This is achieved by
means of morphological processing followed by the circular
Hough transform on the image gradient to detect the contour
of the optic disk within a circular region of interest. The
number of edge pixels and the number of radii used is
reduced by applying Hough transform only to the gradient
image, since the computational complexity of the Hough
transformation is highly dependent on the number of edge
pixels and the number of radii to be matched.
[attachment=23456]
INTRODUCTION
Fundus images are used for diagnosis by trained clinicians
to check for any abnormalities or any change in the retina.
They are captured by using special devices called
ophthalmoscopes. A typical fundus image with its features
marked is shown in the Figure 1. Each pixel in the fundus
image consists of three values namely red, green and blue,
each value being quantised to 256 levels.
The information about the optic disk can be used to
examine severity of some diseases such as glaucoma.
Changes in the optic disk can indicate the current state and
progression of a certain disease. The location of the optic
disk is an important issue in retinal image analysis as it is a
significant landmark feature, and its diameter is usually used
as a reference length for measuring distances and sizes.
METHODOLOGY
The optic disk appears in colour fundus images as a bright
yellowish or white region. Its shape is more or less circular,
interrupted by outgoing vessels. Although sometimes due to
the nature of the photographic projection it has the form of
an ellipse.
Mathematical morphology in image processing is
particularly suitable for analysing shapes in images. The two
main processes are those of dilation and erosion. These
processes involve a special mechanism of combining two
sets of pixels. Usually, one set consists of the image being
processed and the other a smaller set of pixels known as a
structuring element or kernel.
RESULTS
The proposed technique was tested on the DRIVE database
of retinal images [8]. This consists of 40 fundus images of
dimensions 768×584, captured by a Canon CR5 nonmydriatic
3CCD camera with a 45o field of view (FOV).
These images contain both normal (healthy) and abnormal
retinas. In this study 38 of these images were used (2 images
have been excluded for not having visually-detectable optic
disks) consisting of 31 normal and 7 abnormal images. The
performance of the optic disk localisation was evaluated
based on the determined optic disk location with regard to
an expert. Our proposed method is capable of localising the
optic disk correctly for 36 of these images. The success rate
was 94.7%. This was a notable result compared to the
method used in [7] achieving 90.25% for the same DRIVE
dataset.
CONCLUSIONS
In this paper, a computer-assisted retinal image analysis
system for the localisation of optic disk in colour digital
fundus images has been presented. This is achieved by
means of morphological processing followed by the circular
Hough transform on the image gradient to detect the contour
of the optic disk within a circular region of interest. The
number of edge pixels and the number of radii used is
reduced by applying Hough transform only to the gradient
image, since the computational complexity of the Hough
transformation is highly dependent on the number of edge
pixels and the number of radii to be matched.