Skripsi

ANALISIS PENGARUH VARIASI RUANG WARNA PADA FITUR CLAHE TERHADAP KLASIFIKASI PENYAKIT MATA BERBASIS EFFICIENTNET-B0

---

Automatic classification of eye diseases (cataract, glaucoma, and diabetic retinopathy) is often constrained by low contrast and illumination bias in retinal fundus images. This research aims to analyze the effect of the Contrast Limited Adaptive Histogram Equalization (CLAHE) method on color space variations to improve the accuracy of the EfficientNet-B0 deep learning architecture. The evaluation was conducted using 4,217 images from Kaggle which were partitioned into training, validation, and testing data with a ratio of 70:20:10. The experiment compared three pre-processing scenarios: CLAHE on the Green channel (RGB), Luminance channel (CIE Lab), and Luma channel (YCbCr). The testing results prove that the application of CLAHE on the Luma channel (YCbCr) provides the most optimal performance with a Total Accuracy reaching 92.27% and a Macro F1-Score of 92.08%. This figure surpasses the performance of the RGB (89.70%) and CIE Lab (90.87%) color spaces. In conclusion, the separation of light intensity in the YCbCr color space significantly emphasizes retinal pathological features without destroying the original color. Despite the high accuracy, the detection of the glaucoma class remains a challenge that requires a specific optical segmentation approach in future research

---

Availability

Detail Information

Series Title

-

Call Number

T2003162026

Publisher

Indralaya : Prodi Teknik Informatika, Fakultas Ilmu Komputer Universitas Sriwijaya., 2026

Collation

xvi, 123 hlm.; ilus.; tab.; 29 cm.

Language

Indonesia

ISBN/ISSN

-

Classification

617.707

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

penyakit mata
Prodi Teknik Informatika

Specific Detail Info

-

Statement of Responsibility

MI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment