Visualization of exudates fundus images using radar chart and colour auto correlogram technique for early detection of diabetic retinopathy / Hasliza Abu Hassan
According to World Health Organisation (WHO) statistics in 2011, about 175 million people at present are with diabetes mellitus and it will increase to 366 million by 2030. Currently, there are several tests that have been used to early detect diabetes disease and these tests are difficult and hard...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2018
|
| Subjects: | |
| Online Access: | https://ir.uitm.edu.my/id/eprint/82556/1/82556.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | According to World Health Organisation (WHO) statistics in 2011, about 175 million people at present are with diabetes mellitus and it will increase to 366 million by 2030. Currently, there are several tests that have been used to early detect diabetes disease and these tests are difficult and hard on patients and can result in nausea. The alternative way to make it easy on patients is by analyzing the fundus image of retina to early detect or controlling the diabetes disease. Hence in this research, visualization of exudates in fundus images using Color Auto Correlogram (CAC) technique and radar chart for early detection of Diabetic Retinopathy (DR) is proposed. The method consists of capturing several key features of the CAC extracted from fundus images and plotting these features in the form of radar chart. The method is centred on the detection and removal of exudates from the original fundus image, then comparing the CAC features of the original image with the exudate removed from the image. If exudates are present, the CAC exhibits significantly different features which can be easily distinguished using the radar chart. A total of 149 fundus images from the publicly available MESSIDOR databases were used as database in this study. |
|---|