Automatic Detection of Abnormalities in Retinal Blood Vessels using DTCWT, GLCM Feature Extractor and CNN-RNN Classifier
Revathi priya Muthusamy1, S. Vinod2, M. Tholkapiyan3
1Revathi priya Muthusamy, Department of Computer Science and Engineering, Vel Tech Multi Tech Dr.Rangarajan Dr.Sakunthala Engineering College, 42,Alamathi Road,Avadi,Thiruvallur, Chennai, (Tamil Nadu), India.
2S. Vinod, Assistant Professor, Department of Computer Science and Engineering, Vel Tech Multi Tech Dr.Rangarajan, Dr.Sakunthala Engineering College, 42,Alamathi Road,Avadi,Thiruvallur, Chennai, (Tamil Nadu), India.
3Dr. M. Tholkapiyan,Professor, Department of Department of Civil Engineering, Saveetha School of Engineering, Saveetha Institute of Medical and Technical Sciences, Tandalam, Chennai (Tamil Nadu), India.
Manuscript received on 23 March 2019 | Revised Manuscript received on 30 March 2019 | Manuscript published on 30 March 2019 | PP: 1890-1893 | Volume-7 Issue-6, March 2019 | Retrieval Number: F2875037619/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: In worldwide, retinal diseases are found to be frequent cause of blindness for working age population in western countries. So, early diagnosis can prevent the blindness. We develop a system for the early diagnosis of retinal disease. The images with different colour variation inside the eye is compared by using images taken laser camera with high definition. These images are termed as fundus images. Images processing technologies are employed as follows: The feature extraction of the fundus images can be obtained by using the software tool MATLAB. Automatic screening will help to quickly identify the condition of the patients in a more accurate way. The 4-level discrete wavelet transform is used to decompose the image into various sub-bands. The textural features had been calculated using GLCM features, and the classification is done by using CNN-RNN Neural networks. The processed output will be displayed using Matlab GUI. Experimental result proves that the abnormality in the blood vessels and exudates can be effectively detected by applying this method on the retinal images. 76% of test cases are correctly classified.
Keywords: Retinal , Funds image, MATLAB, DTCWT, GLCM, CNN-RNN.
Scope of the Article: Image Security