Glaucoma Detection using CNN and Matlab

Ask For Price


Digital images are obtained from the retina and graded by trained professionals. The progression of glaucomatous retinopathy is assessed by its severity, which in turn determines the frequency of examinations. However, a significant shortage of professional observers has prompted computer-assisted monitoring. In Our Proposed method we implement pre-processing using noise removal or green plane process in Discrete Wavelet Transform, Curvelet Transform, Orthogonal Transform, and Fuzzy segmentation on a minimum of 10 features (Mean, Variance, Entropy?.) with our data set trained images on 50 images using NN (neural network) training and we applying NN classifier based Normal or Abnormal in accuracy comparison between Discrete Wavelet, Curvelet, and Orthogonal transform and fuzzy segmentation. When the blood vessels get damaged Glaucoma disorder is found


Retinal vessel segmentation is the implementation of screening programs for glaucomatous retinopathy, evaluation of the retinopathy of prematurity, detection of macular avascular regions, detection of arteriolar narrowing, measurement of vessel tortuosity to characterize hypertensive retinopathy, measurement of vessel diameter to diagnose hypertension, cardiovascular diseases, and computer-assisted laser surgery. Other indirect applications include automatic generation of retinal maps for the treatment of age-related macular degeneration, extraction of characteristic points of the retinal vasculature for temporal or multimodal image registration, retinal image mosaic synthesis, identification of the optic disc position, and localization of the fovea. Furthermore, the network of retinal vessels is distinctive enough to each individual and can be used for biometric identification, although it has not yet been extensively explored. Vessels, fovea, and optical disks are the three most important structures of the human retina and are mostly used for several applications, such as retinal image registration, illumination correction, as well as pathology detection inside the retina. Detection of these important structures manually is time-consuming and depends on the expertise of the user. The segmentation of blood vessels from fundus photographs can be difficult for a number of reasons. Some of the corrupting sources are related to the acquisition process and kind of imagery, and others are intrinsic features of retinal images


  • DCT


  • In this method, intrinsic characteristics of retinal images make the blood vessel detection process difficult.
  • Poor Edge detection.
  • Less accuracy and poor cluster efficiency.
  • It is not suitable for all lighting conditions images

Proposed System

Retinal image analysis for glaucomatous retinopathy detection based on,

  • Kirsch’s template-based edge Enhancement
  • Morphological filtering
  • Object Classification with Connected component analysis


  • Accurate retina vessel and execute detection
  • Better efficiency and less sensitive to noise
  • It is useful to detect glaucoma in the early stages

Block diagram

GlaucomaDetection using CNN and Matlab

Block Diagram Explanation


Extracting the Green Component from the RGB Image

An RGB image represents each pixel color as a set of three values, representing the red, green, and blue intensities that make up the color. These intensity values are stored directly in the image array. In MATLAB, the red, green, and blue components of an RGB image reside in a single m-by-n-by-3 array. m and n are the numbers of rows and columns of pixels in the image, and the third dimension consists of three planes, containing red, green, and blue intensity values. For each pixel in the image, the red, green, and blue elements combine to create the pixel’s actual color. For further processing, the green component of the RGB input image is extracted.

GlaucomaDetection using CNN and Matlab
?(a) Normal Retina image? ?(b) Green component extracted? image



Wavelet Transform is a type of signal representation that can give the frequency content of the signal at a particular instant of time or spatial location. Stationary wavelet transform decomposes the image into different subband images, It splits components into numerous frequency bands called subbands. They are? LL, LH, HL, and HH subbands. A high-frequency subband contains the edge information of the input image and an LL subband contains clear information about the image.? Upsampling of coefficients will be performed for factor 2 after each row and column decomposition process to make a lossless representation of edge detail

3. GLCM Features

In statistical texture analysis, texture features are computed from the statistical distribution of observed combinations of intensities at specified positions relative to each other in the image. According to the number of intensity points (pixels) in each combination, statistics are classified into first-order, second-order, and higher-order statistics.

The Gray Level Cooccurrence Matrix (GLCM) method is a way of extracting second-order statistical texture features. The approach has been used in a number of applications, Third and higher-order textures consider the relationships among three or more pixels. These are theoretically possible but not commonly implemented due to calculation time and interpretation difficulty

4. Neural Networks as Classifiers

A neural network consists of units (neurons), arranged in layers, which convert an input vector into some output.? Each unit takes an input, applies an (often nonlinear) function to it, and then passes the output on to the next layer.? Generally the networks are defined to be feed-forward: a unit feeds its output to all the units on the next layer, but there is no feedback to the previous layer.? Weightings are applied to the signals passing from one unit to another, and it is these weightings that are tuned in the training phase to adapt a neural network to the particular problem at hand.? This is the learning phase.

Neural networks have found application in a wide variety of problems.? These range from function representation to pattern recognition, which is what we will consider here.

Backpropagation Neural Network

Artificial neural networks (ANN) consider classification as one of the most dynamic research and application areas. The major disadvantage in using ANN is finding the most appropriate grouping of training, learning, and transfer functions for classifying the data sets with a growing number of features and classified sets. The different combinations of functions and their effect while using ANN as a classifier are studied and the correctness of these functions is analyzed for various kinds of datasets.

The real-world problems which are represented by multidimensional datasets are taken from a medical background. The classification and clustering of these data sets are significant. The data set is divided into a training set and a testing set and it has no usage in the training process. The results are produced with the help of these datasets and it is used for testing. The training set is taken from 2/3rd of the dataset and the remaining has been taken as a test set. This is made through the assessment of the accuracy achieved through testing against these data sets. Then the network is simulated with the same data.

The backpropagation algorithm trains the neural network. The gradient descent method (GDM) was used to decrease the mean squared error between the network output and the actual error rate. The following parameters are considered to measure the efficiency of the network,

Rate of convergence

  • No epochs were taken to converge the network.
  • The calculated Mean Square Error (MSE)


Morphological image processing is a collection of non-linear operations related to the shape or morphology of features in an image. Morphological operations rely only on the relative ordering of pixel values, not on their numerical values, and therefore are especially suited to the processing of binary images. Morphological operations can also be applied to greyscale images such that their light transfer functions are unknown and therefore their absolute pixel values are of no or minor interest.

Morphological techniques probe an image with a small shape or template called a structuring element. The structuring element is positioned at all possible locations in the image and it is compared with the corresponding neighborhood of pixels. Some operations test whether the element “fits” within the neighborhood, while others test whether it “hits” or intersects the neighborhood: Probing of an image with a structuring element

A morphological operation on a binary image creates a new binary image in which the pixel has a non-zero value only if the test is successful at that location in the input image.

The structuring element is a small binary image, i.e. a small matrix of pixels, each with a value of zero or one:

  • The matrix dimensions specify the size of the structuring element.
  • The pattern of ones and zeros specifies the shape of the structuring element.
  • An origin of the structuring element is usually one of its pixels, although generally, the origin can be outside the structuring element.

Requirement Specification

Software Requirement

  • python above 3.0
  • Neural network Toolbox

Hardware Requirements

  • system
  • 4 GB of RAM
  • 500 GB of Hard disk


[1] M. E. Martinez-Perez, A. D. Hughes, S. A. Thom, and K. H. Parker,?Improvement of a retinal blood vessel segmentation method using the insight segmentation and registration toolkit (ITK),? in Proc. IEEE 29th Annu. Int. Conf. EMBS. Lyon, IA, France, 2007, pp. 892? 895.

[2] S. Dua, N. Kandiraju, and H.W. Thompson, ?Design and implementation of a unique blood-vessel detection algorithm towards early diagnosis of glaucomatous retinopathy,? in Proc. IEEE Int. Conf. in Inf. Technol., Coding Comput., 2005, pp. 26? 31.

[3] A.M.Mendonc?a and A. Campilho, ?Segmentation of retinal blood vessels by combining the detection of centerlines and morphological reconstruction,?IEEE Trans. Med. Imag., vol. 25, no. 9, pp. 1200?1213, Sep. 2006.

[4] A. Aquino, M. E. Geg?ndez-Arias, and D. Mar?n, ?Detecting the optic disc boundary in digital fundus images using morphological, edge detection, and feature extraction techniques,? IEEE Trans. Med. Imag., vol. 29, no. 11, pp. 1860?1869, Nov. 2010.

[5] M. Lalonde, M. Beaulieu, and L. Gagnon, ?Fast and robust optic disc detection using pyramidal decomposition and Hausdorff-based template matching,? IEEE Trans. Med. Imag., vol. 20, no. 11, pp. 1193?1200, Nov. 2001

[6] T.Kauppi and H.K?lvi?inen, ?Simple and robust optic disc localization using colour decorrelated templates,? in Proc. 10th Int. Conf. Advanced Concepts for Intell. Vision Syst.,. Berlin, Germany: Springer-Verlag, 2008, pp. 719?729.


Customer Reviews

There are no reviews yet.

Be the first to review “Glaucoma Detection using CNN and Matlab”

Your email address will not be published. Required fields are marked *