Convolution neural network model for fundus photograph quality assessment

• DOI: 10.11591/ijeecs.v26.i2.pp915-923
• Publication Date: 2022-05-02T15:40:29Z
• Authors: Sinan S. Mohammed...; Tian-Swee Tan; Muhammad Amir Bin...; Wan Hazabbah Wan ...; Qi Zhe Ngoo; Joyce S.Y. Sia; Kelvin Ling Chia ...
• Keywords: Fundus (uterus); Convolution (computer science)

<span>The excellent quality of color fundus photograph is crucial for the ophthalmologist to process correct diagnosis and convolutional neural network (CNN) models optimize output classification. As a result main causes as acquire devises efficiency experience physician most photographs can have uneven illuminance, blur, bad contrast, in addition micro-features retinal diseases, which need force their contrast. Fundus assessment method proposed find out perfect enhanced Technique fundoscopy photographs-based CNN model. Five measurements, five metrics, were used standard this study. In research innovative approach combining measurement metrics analysis best enhance that set multiclass The contrast enhancement techniques are evaluated using 267 divided into three retina diseases cases downloaded from open-source database “FIGSHARE”. study outcome showed presented system (single-CNN) determine well method, low-quality then it boost achieve superior.</span>