AUTOMATIC DETECTION AND DIAGNOSIS OF

DIABETIC RETINOPATHY

1.1 INTRODUCTION
In recent times, India and other parts of the world have been faced with an increase in
age and society related diseases like diabetes. According to recent survey, 24% of the country
population has been diagnosed of diabetes disease alone and it have been recognize and accepted
as one of the main cause of blindness in the country if not properly treated and managed. Early
detection and diagnosis have been identified as one of the way to achieve a reduction in the
percentage of visual impairment caused by diabetes with more emphasis on routine medical
check which the use of special facilities for detection and monitoring of the diabetes.
Diabetic related eye diseases are the most common cause of blindness in the world.
Diabetic Retinopathy is a severe and widely spread eye disease, which can be regarded as
manifestation of diabetes on retina. Diabetic Retinopathy is a specific micro vascular
complication of both insulin dependant(type 1) and non insulin dependant(type 2) diabetes. The
prevalence of retinopathy s strongly linked to the duration of diabetes. After 20 years of diabetes
nearly all patients with type 1 diabetes and over 60% of patients with type 2 diabetes have some
degree of retinopathy.
Vision losses often, late symptoms of advanced diabetic retinopathy, many patients
remain undiagnosed even as their disease is causing severe retinal damage. Hence there is an
urgent need for mass screening retinal examination for the early detection and treatment of
diabetic retinopathy.

DIABETIC RETINOPATHY
ANATOMY OF HUMAN EYE
The eye is small, only about 1 inch in diameter, it serves a very important function the
sense of sight..The eye is often compared to a camera. Each gathers light and then transforms
that light into a picture. Both also have lenses to focus the incoming light. A camera uses the
film to create a picture, whereas the eye uses a specialized layer of cells, called the retina, to
produce an image.
The orbit is the eye socket, The eye is cushioned within the orbit by pads of fat. The orbit
also contains the lacrimal gland that is located underneath the outer portion of the upper eyelid.
The eyelids serve to protect the eye from foreign matter, such as dust, dirt, and other debris, as
well as bright light that might damage the eye. On Blinking the eyelids also help spread tears
over the surface of your eye, keeping the eye moist and comfortable. The eyelashes help filter
out foreign matter, including dust and debris, and prevent it from getting into the eye. The
conjunctiva is a thin, clear layer of skin covering the front of the eye, including the sclera and the
inside of the eyelids. The conjunctiva keeps bacteria and foreign material from getting behind the
eye.

Figure 1.1.1 Anatomy of human eye

Sclera
The white part of the eye is the front part of the sclera. However, the sclera, a tough,
leather-like tissue, also extends around the eye. The sclera surrounds the eye and gives the eye its
shape. The sclera is also attached to the extraocular muscles, which, in turn, move the eye left
and right, up and down, and diagonally.
Cornea
The cornea is a clear layer at the front and center of the eye. In fact, the cornea is so clear
that you may not even realize it is there. The cornea is located just in front of the iris, which is
the colored part of your eye. The main purpose of the cornea is to help focus light as it enters the
eye. If you wear contact lenses, the contact lens rests on your cornea.
Choroid

The choroid layer is located behind the retina and absorbs unused radiation
Fovea
The fovea is a small depression (approx. 1.5 mm in diameter) in the retina.
This is the part of the retina in which high-resolution vision of fine detail is possible.
Iris
The iris is a diaphragm of variable size whose function is to adjust the size of the pupil to
regulate the amount of light admitted into the eye.The iris is the coloured part of the eye
Pupil
The pupil is the aperture through which light passes and hence the images we "see" and
"perceive" - enters the eye. This is formed by the iris. As the size of the iris increases (or
decreases) the size of the pupil decreases (or increases) correspondingly.

Retina
The retina may be described as the "screen" on which an image is formed by light that
has passed into the eye via the cornea, aqueous humor, pupil, lens, then the hyaloid and finally
the vitreous humor before reaching the retina. The retina contains photosensitive elements
(called rods and cones) that convert the light they detect into nerve impulses that are then sent
onto the brain along the optic nerve.
Optic nerve
The

optic

nerve

is

the

second

cranial

nerve

and

is

responsible

for

vision.

Each nerve contains approx. one million fibres transmitting information from the rod and cone
cells of the retina.
5

Ciliary muscles
The ciliary muscle is a ring-shaped muscle attached to the iris. It is important because
contraction and relaxation of the ciliary muscle controls the shape of the lens.
Hyaloid
The hyaloid diaphragm divides the aqueous humor from the vitreous humor.
Aqueous Humor
The aqueous humor is a jelly-like substance located in the anterior chamber of the eye.
Vitreous Humor
The clear, gelatinous substance filling the central cavity of the eye
Visual Axis
A simple definition of the visual axis is a straight line that passes through both the centre of the
pupil and the centre of the fovea.

WORKING OF HUMAN EYE

After light passes through the aqueous humor, it passes through the pupil. This is the
central circular opening in the colored part of the eye -- also called the iris. Depending on how
much light there is, the iris may contract or dilate, limiting or increasing the amount of light that
gets deeper into the eye. The light then goes through the lens. Just like the lens of a camera, the
lens of the eye focuses the light. The lens changes shape to focus on light reflecting from near or
distant objects.This focused light now beams through the center of the eye. Again the light is
bathed in moisture, this time in a clear jelly known as the vitreous. Surrounding the vitreous is
the retina.Light reaches its final destination within the photo receptors of the retina: The light
focuses on the retina, a series of light-sensitive cells lining the back of the eye. The retina acts
like camera film, reacting to the incoming light and sending a record of it via the optic nerve to
brain
6

DIABETES AND DIABETIC RETINOPATHY

Diabetes is a disorder of metabolism. The energy required by the body is obtained from
glucose which is produced as a result of food digestion. Digested food enters the body stream
with the aid of a hormone called insulin which is produced by the pancreas, an organ that lies
near the stomach. During eating, the pancreas automatically produces the correct amount of
insulin needed for allowing glucose absorption from the blood into the cells. In individuals with
diabetes, the pancreas either produces too little or no insulin or the cells do not react properly to
the insulin that is produced. The build up of glucose in the blood, overflows into the urine and
then passes out of the body. Therefore, the body loses its main source of fuel even though the
blood contains large amounts of glucose.
Diabetic retinopathy is a disorder of the RV arising from acute diabetic mellitus. It is
caused by changes in the blood vessels of the retina. In some people with diabetic retinopathy,
blood vessels may swell and leak fluid. In other people, abnormal new blood vessels grow on the
surface of the retina. The retina is the light-sensitive tissue at the back of the eye. A healthy
retina is necessary for good vision. If you have diabetic retinopathy, at first you may not notice
changes to your vision. But over time, diabetic retinopathy can get worse and cause vision loss.
Diabetic retinopathy usually affects both eyes.

STAGES OF DIABETIC RETINOPATHY

Mild Nonproliferative Retinopathy. At this earliest stage, microaneurysms occur. They are
small areas of balloon-like swelling in the retina's tiny blood vessels.
Moderate Nonproliferative Retinopathy. As the disease progresses, some blood vessels that
nourish the retina are blocked.
Severe Nonproliferative Retinopathy. Many more blood vessels are blocked, depriving
several areas of the retina with their blood supply. These areas of the retina send signals to the
body to grow new blood vessels for nourishment.
7

Proliferative Retinopathy. At this advanced stage, the signals sent by the retina for

nourishment trigger the growth of new blood vessels. This condition is called proliferative
retinopathy. These new blood vessels are abnormal and fragile. They grow along the retina and
along the surface of the clear, vitreous gel that fills the inside of the eye. By themselves, these
blood vessels do not cause symptoms or vision loss. However, they have thin, fragile walls. If
they leak blood, severe vision loss and even blindness can result.

SIGNS AND SYMPTOMS

The various symptoms of diabetic retinopathy are
Difficulty in reading
Blurred vision
Sudden loss of vision in one eye
Seeing rings around lights
Dark spots or flashing lights
Pregnancy and high blood pressure may aggravate diabetic retinopathy.
The clinical manifestation of Diabetic retinopathy is exudates and hemorrhages.

Hemorrhages: They occur in the deeper layers of the retina and are often called blot
hemorrhages, because of their round shape. As new blood vessels form at the back of the
eye as a part of proliferative diabetic retinopathy (PDR), they can bleed (ocular
hemorrhage) and blur vision. In most cases, it will leave just a few specks of blood, or
spots, floating in a person's visual field, though the spots often go away after a few
hours.These spots are often followed within a few days or weeks by a much greater
leakage of blood, which blurs vision.

Exudates: They are one of the main characteristics of diabetic retinopathy and can vary
in size from tiny specks to large patches with clear edges. As well as blood, fluid that is
rich in fat and protein is contained in the eye and this is what leaks out to form the
exudates. These can impair vision by preventing light from reaching the retina.
Currently the exudates of diabetic retinopathy are described as small ,sharply
demarcated yellow or white, waxy glistering patches ,often coalescing in to plaques.
These exudates are usually termed waxy exudates or hard exudates because they
have an appearance of hardness.

Microaneurysms (MAs): They are likely to be the only lesion present at the earliest
stage and during the process of disease development.. MAs are swellings of the
capillaries caused by a weakening of the vessel wall. In retinal photographs, although the
capillaries are not visible, MAs appear as dark red isolated dots. Microaneurysms is the
earliest clinical sign of diabetic retinopathy. They appear as small,dots in the superficial
retinal layers.

10

DIAGNOSIS
Diabetic retinopathy is detected during an eye examination that includes:

Visual acuity test: This test uses an eye chart to measure how well a person sees at
various distances (i.e., visual acuity).

Pupil dilation: The eye care professional places drops into the eye to widen the pupil.
This allows him or her to see more of the retina and look for signs of diabetic retinopathy.
After the examination, close-up vision may remain blurred for several hours.

Ophthalmoscopy: This is an examination of the retina in which the eye care professional:
looks through a device with a special magnifying lens that provides a narrow view of the
retina, or wearing a headset with a bright light, looks through a special magnifying glass
and gains a wide view of the retina. Hand-held ophthalmoscopy is insufficient to rule out
significant and treatable diabetic retinopathy.

Optical coherence tomography (OCT): This is an optical imaging modality based upon
interference, and analogous to ultrasound. It produces cross-sectional images of the retina
11

(B-scans) which can be used to measure the thickness of the retina and to resolve its
major layers, allowing the observation of swelling and or leakage.

Digital Retinal Screening Programs: This involves digital image capture and transmission
of the images to a digital reading center for evaluation and treatment referral.

Slit Lamp Biomicroscopy Retinal Screening Programs: Systematic programs for the early
detection of diabetic retinopathy using slit-lamp biomicroscopy.

TREATMENT
During the first three stages of diabetic retinopathy, no treatment is needed, unless you
have macular edema. To prevent progression of diabetic retinopathy, people with diabetes should
control their levels of blood sugar, blood pressure, and blood cholesterol.
Proliferative retinopathy is treated with laser surgery. This procedure is called scatter
laser treatment. Scatter laser treatment helps to shrink the abnormal blood vessels. Doctor places
1,000 to 2,000 laser burns in the areas of the retina away from the macula, causing the abnormal
blood vessels to shrink. Because a high number of laser burns are necessary, two or more
sessions usually are required to complete treatment. Scatter laser treatment may slightly reduce
color vision and night vision.
Scatter laser treatment works better before the fragile, new blood vessels have started to
bleed. That is why it is important to have regular, comprehensive dilated eye exams. Even if
bleeding has started, scatter laser treatment may still be possible, depending on the amount of
bleeding. If the bleeding is severe, you may need a surgical procedure called a vitrectomy.
During a vitrectomy, blood is removed from the center of your eye.

12

SCOPE AND SIGNIFICANCE OF THE PROJECT

An ophthalmologist obtain retinal images from the fundus camera for the patients to be
diagnosed. From the image symptoms will be identified manually by an ophthalmologist,
therefore the more patients to be diagnosed, the more time will be needed. A computerized
screening system can be used for fully automated mass screening. Such systems screen a large
number of retinal images and identify abnormal images, which are then further examined by an
ophthalmologist. This would save a significant amount of workload and time for
ophthalmologists, allowing them to concentrate their resources on surgery and treatment.
Computerized system can also be used for extracting the other features apart from the three
major symptoms. The software can also be implemented in the hospitals to detect diabetic
retinopathy thereby reducing manual intervention. The information provided by the system also
helps the doctors in the retinal surgery. This study can be extended for the analysis of other
diseases like hypertension, stroke, migraine, hearing loss etc. Methods are currently being
investigated for improving performance

LITERATURE SURVEY
Automated Detection and Classification of Vascular abnormalities in
Diabetic Retinopathy
Vallabha, D. Dorairaj, R. Namuduri, K. Thompson, H. Dept. of Electr. & Comput. Eng.,,
Wichita State

Univ., KS, USA Date of Current Version: 21 March 2005

Use of scale and orientation of selective Gabor filter to detect and classify the retina
images into mild or severe case. Has ability to distinguish images by virtues of its variation
across scales and orientation. The input image is first filtered through Gabor filter banks. The
banks consist of several filters tuned to specific scales and orientation and the operation - Fourier
domain. Vascular abnormalities are efficiently captured in the finer scales of the Gabor filter

13

outputs. These abnormalities appear at high frequencies but they do not appear in the low-pass
filter outputs.
The output is analyzed. Detection of PDR is done by analyzing the width of the blood
vessels. The presence of one local maxima in the plot of energy vs. orientation for more than 100
test images signify the presence of mild to PDR while the presence of more than one local
maxima signify severe PDR.
Disadvantages:

This method only signifies the presence of BDR and PDR but does not specify the
coordinates nor the actual spots or actual disease type.

The specificity and sensitivity of this method were not discussed in work done nor do
they use a full scale image, instead part of the images of size 256 x 256 pixels were used

II A Development of Computer Aided Diagnosis (CAD) system using

fundus images
Hayashi, J.; Kunieda, T.; Cole, J.; Soga, R.; Hatanaka, Y.; Miao Lu; Hara, T.; Fujita,
H.; Softopia Japan, Ogaki .Date of Current Version: 06 August 2002.
A CAD system can help physicians by displaying useful information such as the location
of abnormalities. Although many kinds of images are used in medical diagnosis, we have
focused our research to methods of analyzing fundus images. Physicians normally observe the
condition of blood vessels and the retina when examining fundus images. The initial four steps
involve converting and RGB fundus image to monochrome; from this areas of low density were
14

extracted using Binarization method. Followed by deletion of vascular regions and the last stage
involves deletion of unnecessary elements. The method listed above suffered from lots of
misdetection and it necessitate improvement. In detecting white spots related abnormalities, the
same method applied for the red spots were used on the negative of the monochrome image.
Two detection methods yield misdetection. Improvement method suggested on
misdetection around the optic disk offers improvement only for fundus images with visible optic
disk .Misdetection that occurred with multiple blood vessels were left unresolved.

III

Automatic Detection of Red lesions from Color Fundus Photographs

Michael D. Abrmoff, MD, PHD, Meindert Niemeijer, PHD,3 Maria S.A. Suttorp-Schulten, MD,
PHD, Max A. Viergever, PHD, Stephen R. Russell, MD,and Bram van Ginneken, PHD .Date of
Current Version: 02 May 2005
A novel red lesion detection method is presented based on a hybrid approach, combining
prior works by Spencer and Frame with two important new contributions. The first contribution
is a new red lesion candidate detection system based on pixel classification. Using this technique,
vasculature and red lesions are separated from the background of the image. After removal of the
connected vasculature the remaining objects are considered possible red lesions. Second, an
extensive number of new features are added to those proposed by SpencerFrame.
To prepare the fundus images for red lesion extraction some image preprocessing is
performed on the green-plane of the original RGB color image. As red lesions have the highest
contrast with the background in the green color plane, information from the red and blue color
planes is not used in this step .Slow gradients in the background of Igreen were removed,
resulting in a shade corrected image Isc. This was accomplished by estimating the background
image Ibg and subtracting that from Igreen .Ibg is produced by median filtering Igreen with a
15

25*25 pixel kernel. Shade corrected image Isc has negative values for all pixels which have an
intensity lower than the background. By removing all pixels with a positive value from the
image, bright pathologies no longer influence the later analysis. Objects that are potential red
lesions are extracted from the preprocessed images. These objects will be called candidate
objects. To discriminate between circular, nonconnected red lesions and the elongated
vasculature a morphological top-hat transformation was used. This operation is based on
morphologically opening the image with a linear structuring element at different orientations.
When the length of the structuring element is increased to be able to detect larger objects, the
vessel segmentation deteriorates leading to more spurious candidate objects being detected on
the vasculature. We have largely removed this limitation by using a pixel classification based
method. This approach can detect larger candidate objects and reduces the number of spurious
candidate objects on the vessels by integrating a vessel segmentation technique with red lesion
detection. Both vasculature and possible red lesions are extracted at once, and subsequently the
vasculature is separated from the red lesion candidate objects. The detected candidate objects are
classified using all features and a k-nearest neighbor classifier. When determining whether an
image contains red lesions the system achieves a sensitivity of 100% at a specificity of 87%. The
method is compared with several different automatic systems and is shown to outperform them
all. Performance is close to that of a human expert examining the images for the presence of red
lesions

Microaneurysm detection using an optimal wavelet transform

Gwenole Quellec1,3, Mathieu Lamard2,3, Guy Cazuguel1,3, Beatrice Cochener2,3,4,
Christian Roux1,3 INSTITUT TELECOM; TELECOM BretagneUEB; Dpt ITI, Brest, F-29200
France; Univ Bretagne Occidentale, Brest, F-29200 France;Inserm, U650, IFR 148 ScInBioS Science et Ingenierie en Biologie-Sante, Brest, F-29200 France; CHU Brest, Service
dOphtalmologie, Brest, F-29200 France;July 23, 2008
Detect them by locally matching a lesion template in sub bands of wavelet transformed images.
A 2-d model of micro aneurysm is created using Gaussian function called as image model.

16

To perform template-matching in an image I, we move a window w over I and consider

the subimages of I defined by the positions of w, noted I|w. To classify the content of I|w, we
have to find the set of parameters (in our case: , , ,) for which the parametric model best
matches I|w. If the model is close enough to I|w, a lesion is detected at the center of w. The
distance measure commonly used is the sum of the squared errors (SSE): it is defined as the sum
of the squared differences between the value of each pixel of I|w and the corresponding value of
the model function.The position where the best score is obtained is taken as microaneurysm.
Images present brightness variability, local background intensity variations and noise that would
make it difficult to efficiently fit this model to a microaneurysm directly in the spatial domain.
We get round this problem by fitting the wavelet transform of the model (WTM) on the wavelet
transform of the image instead of fitting the original model on the raw image, and considering
only interesting subbands. Indeed, by ignoring the high frequency subbands, we get rid of the
noise, and ignoring the low frequency subbands, we get rid of slow image variations.The best
fitted wavelet, the Haar wavelet, has the most compact support. The results show that the method
efficiency clearly depends on both the wavelet used and the decomposition level. Sometimes
small hemorrhages are frequently detected as microaneurysms.

WAVELET BASED EXTRACTION OF DIABETIC RETINOPATHY

OBJECTIVE
To develop an automated system to analyze the retinal images for important features of
diabetic retinopathy using image processing techniques, wavelet transform and an image
classifier based on artificial neural network which classify the images according to the disease
conditions

17

FEATURES EXTRACTING:

Optic disk

Blood vessels and hemorrhage using wavelet transform

Exudates and Micro aneurysm using imaging techniques and wavelet transform

CLASSIFIES RETINAL IMAGES AS:

No DR(Normal retina)

Hemorrhage image.

Microaneurysm image

Exudate image

BLOCK DIAGRAM
The retinal images are obtained as RGB images.These RGB images are preprocessed using
the image processing techniques like gray scale conversion and morphological operations like
closing and opening and the main three symptoms of diabetic retinopathy like hemorrhage,
microaneurysm ,exudates are exracted using wavelet transform.The artificial neural network
using back propagation neyral network classifies it according to the the presence of the symptom

18

PRE PROCESSING

RGB IMAGE

EXTRACTING
FEATURE OF
DIABETIC
RETINOPATHY
NEURAL
NETWORK
CLASSIFIER

RESULT

DETECTION OF AN OPTIC DISK

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. 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

BLOCK DIAGRAM
The RGB image is converted to gray scale histogram equalization is performed
and the filtered image undergo closing followed by opening

RGB
IMAGE

GRAY
SCALE
IMAGE

MEDIAN
FILTRING

OPTIC DISK

IMAGE
OPENING

IMAGE
CLOSING

Fig : 3.3.1 Preprocessing steps

19

BLACK &
WHITE
IMAGE

DETECTION OF BLOOD VESSEL

Blood vessels can be described as dark curvilinear objects set against a lighter
background with ill defined edges. The retinal blood vessels are non uniform in intensity, length
and width throughout the image

CONVOLUTION
Convolution is a mathematical operation on two functions f and g, producing a third function
that is typically viewed as a modified version of one of the original functions. Convolution is
similar to cross-correlation. In image processing the convolution operator is used as a filter to
change the characteristics of the image; sharpen the edges, blur the image or remove the high or
low frequency noise. In seismic processing a convolution can be used to extrapolate the
propagating wavefield forward of backward. In signal processing it can be used to suppress
unwanted portions of the signal or separate the signal in different parts.
Two dimensional (2D) convolutions are sometimes the most time consuming parts of an
application.. A Gaussian blur (also known as Gaussian smoothing) is the result of blurring an
image by a Gaussian function or applying a Gaussian blur to an image is the same
as convolving the image with a Gaussian function; It is a widely used typically to reduce image
noise and reduce detail. Applying a Gaussian blur has the effect of reducing the image's highfrequency components; a Gaussian blur is thus a low pass filter. The blurred image is then
subtracted from original gray scale image to detect the blood vessel.

DETECTION OF MICROANEURYSM
Microaneurysm are the symptoms of diabetic retinopathy where the bright lesions or red
colour spots appear in the eye.
20

A model for microaneurysms

Despite their size and intensity variations, microaneurysms (MAs) are quite similar to
each other. Indeed, we can model them with 2-dimensional rotation-symmetric generalized
Gaussian functions, defined by the following equation

??????????? ? ? ??????????????

?? ? ?? ? ??

is the parameter modeling lesion size

is the parameter modeling lesion sharpness: it is a shape factor

is the parameter modeling the background intensity

is the parameter modeling lesion height

The wavelet transform

Wavelets are families of functions generated from one single prototype function (mother
wavelet) by dilation and translation operations: The mother wavelet is constructed from the socalled scaling function, satisfying the two-scale difference equation

(t)= ??

??

???????

??

where h(k) are the wavelet coefficients. Then, the mother wavelet (t) is defined as

(t)= ??

??

???????

??

where ,
g(k) = (1)kh(1 k) .
Several different sets of coefficients h(k) can be found, which are used to build a unique and
orthonormal wavelet basis . The wavelet transform represents the decomposition of a function
into a family of wavelet functions m,n(t) (where m is the scale/dilation index and n the
21

time/space index).. Many constructions of wavelets has been introduced in the mathematical and
in the signal processing literature . The wavelet transform may be seen as a filter bank and
illustrated as follows, on a one dimensional signal x[n]:
x[n] is high-pass and low-pass filtered, producing two signals d[n] (detail) and c[n]
(coarse approximation)
d[n] and c[n] may be subsampled (decimated by 2: 2),otherwise the transform is called
translation invariant wavelet transform
The process is iterated on the low-pass signal c[n]

DETECTION OF EXUDATES
Exudates are a visible sign of diabetic retinopathy and also a marker for the presence of
co-existent retinal edema. Exudates can be of two types; Hard or Soft. Hard exudates are
accumulated lipid formations leaked from weakened vessels. Hard exudate lesions are waxy and
yellow with relatively clear edges, and often appear in clusters or rings. Soft exudates, also called
cotton wool spots or micro-infarctions, appear when terminal retinal arterioles are obstructed.
Soft exudates are small, whitish lesions with blurry edges

3.6.1 Morphological Image Processing

The major part of morphological operations can be defined as a combination of two
basic operations, dilation and erosion, and non-morphological operations like difference, sum,
maximum or minimum of two images. Morphological operations also make use of a structuring
element M; which can be either a set or a function that correspond to a neighborhood function
related to the image function g(x). Further morphological operations and algorithms can be
obtained from sequencing the basic operations. In general, a dilation is every operator that
commutes with the supremum operation. On the other hand, an erosion (denoted by is every
operator that commutes with the infimum operation. There is a homomorphism between the
image function G and the set B of all pixels with image function value 1. The structuring element
is a function that assigns a subset of N x N to every pixel of the image function. Then dilation, an
increasing transformation is defined as

22

Similarly, opening of set by structuring element B is defined as

and closing of set by B structuring element M is defined as

OPERATION

1. Image acquisitions
2. Grey scale conversion
3. Binary image conversion
4. Feature extraction
5. Blood clotted area calculation
6. Vein map and large vein size detection
7. Extraction of image of optical disk
8. Lesions detection

23

Image acquisitions
Image acquisition in image processing can be broadly defined as the action of
retrieving an image from some source, usually a hardware-based source, so it can
be passed through whatever processes need to occur afterward. Performing image
acquisition in image processing is always the first step in the workflow sequence
because, without an image, no processing is possible.
One of the forms of image acquisition in image processing is known as real-time
image acquisition. This usually involves retrieving images from a source that is
automatically capturing images. Real-time image acquisition creates a stream of
files that can be automatically processed, queued for later work, or stitched into a
single media format. One common technology that is used with real-time image
processing is known as background image acquisition, which describes both
software and hardware that can quickly preserve the images flooding into a system.
This is the image acquisition block in LABVIEW

24

STEPS IN IMAGE ACQUISITION

1. First we have captured the image of human eye. That image is captured by a
special type of camera known as fundus camera. This camera is taking a RGB
image of eye.
2. Then the image is converted in to unsigned 32 bit images.
3. Fundus photography is the creation of a photograph of the interior surface of
the eye, including the retina, optic disc, macula , and posterior pole
Fundus photography is used by optometrists, ophthalmologists, and trained
medical professionals for monitoring progression of a disease, or in screening
programs and epidemiology.
4. Then the output images are going for further processing. That image is going for
grey scale conversion.
Grey scale conversion
A gray scale digital image is an image in which the value of each pixel is a
single sample, that is, it carries only intensity information. Images of this sort, also
known as black-and-white, are composed exclusively of shades of gray, varying
from black at the weakest intensity to white at the strongest.
Gray scale images are distinct from one-bit bi-tonal black-and-white images,
which in the context of computer imaging are images with only the
two colors, black, and white (also called bi-level or binary images). Gray scale
images have many shades of gray in between. Gray scale images are also
called monochromatic, denoting the presence of only one (mono) color (chrome).

25

STEPS IN GRAY SCALE CONVERSION

1. The image from the image acquisition comes to gray scale conversion.
2. Then the image comes to geometry block. In geometry block we can do
resampling, resize and rotation.
3. But here we have done resampling.
4. After resampling we have done color plane extraction. We have extracted
the green plane from the image.
5. Extracted image fed to the binary image conversion system.
BINARY IMAGE CONVERSION
A binary image is a digital image that has only two possible values for
each pixel. Typically the two colors used for a binary image are black and white
though any two colors can be used. The color used for the object(s) in the image is
the foreground color while the rest of the image is the background color. In the
document-scanning industry this is often referred to as "bi-tonal".
Binary images are also called bi-level or two-level. This means that each pixel is
stored as a single biti.e., a 0 or 1. The names black-and-white, B&W,
26

monochrome or monochromatic are often used for this concept, but may also
designate any images that have only one sample per pixel, such as gray scale
images.

BLOOD CLOTTED AREA CALCULATION

We have performed XOR operation on the given two extracted image, then simply
converted the final image to the 2-D array and summed up using summer. The summer
output gives the magnitude of the blood clotted area.we have then taken a comparator
to decide whether the given test image comes under Hemorrage, Exduates e.t.c

27

1.image

2.image

3.image

1.After performing boundary rejection and homomorphic operation image.1 is obtained

2.After performing averaging operation we have obtained the image 2
3.Finally 3.image = 1.image XOR 2.image

OPTICAL DISK EXTRACTION

Optic disc or optic nerve (ON) head extraction in retinal images has widespread
applications in retinal disease diagnosis and human identification in biometric systems.
This paper introduces a fast and automatic algorithm for detecting and extracting the
28

ON region accurately from the retinal images without the use of the blood-vessel
information.

STEPS IN OPTICAL DISK EXTRACTION

1. In optical disk extraction first we have done the color plane extraction. We have
to extract the green components from the image of human eye.
2. Then we have resampled the image.
3. After the resampling we have done image smoothing. Image smoothing is done
by median filter. The sharp edges which have contain high frequencies extracted
by low pass filtrer.

4. After smoothing the image we passed the image through a comparator whose
Pixel values intensity >197 change to Red.

29

LESIONS DETECION
The detection of lesions facilitates in initial screening step of the disease, with this
we can perform automatic screening of images whether they are DR infected or not. In
present system with the help of morphological image processing techniques, we are
trying to detect lesions in two categories i.e. dark and bright lesions. The present
system is able to detect 90 % exudates in image and 85% dark lesions.
In Lesions detection we have to find out the number of swelling veins in eye.
According to swelling veins we are categorizing the diseases into three categories.
1. Microneurosyms
2. Hemorrhages
3. Exudates

Blood vessels can be described as dark curvilinear objects set against a lighter
background with ill defined edges. The retinal blood vessels are nonuniform in intensity,
length and width throughout the image

30

SIMULATION RESULTS
Test images taken are of Normal category, Hemorrage stage,Exduates stage and
Microneurosyms stage.
The standered test images are

Hemorrages

Exudates

31

Normal eye

Microaneurosyms

32

LABVIEW SIMULATION RESULT

NORMAL EYE IMAGE

33

When normal image is taken as it is clearly shown that we have observed comparatively
lesser amount of blood clotted area .Also we have only 2 red lesions detected. For
pictorial view the image of optical disk is also clearly visible

Hemorrage affected Eye Image

34

When we have taken image affected with hemorrage we have observed greater
amount of blood clotted area. Also 15 red lesions were detected.Optical disk is
also not clearly visible. Negative retinal image give also the idea that greater
amount of swelling is present in vein.
Exudate image

When we have taken image affected with Hard Exduates we have also
comparitevely greater amount of blood clotted area. Number of red lesions
detected is also 13 which is greater than normal eye but slightly less than
Hemorrage affected eye. Optical disk is also not clearly visible.

35

COMPARISON TABLE

Area

Redlesion

Normal
66

Hemorage
8295

15

Exudates
5121

36

13

CONCLUSION

Detection of optic disk, exudates, microaneurysm and hemorrhages was possible,

with success rates dependent upon preprocessing and the number of images used in training.
When compared with the ophthalmologist, the network achieved good accuracy for the detection
of diabetic retinopathy. The system could be used as an aid to the screening of diabetic patients
for retinopathy.

37