SEGMENTASI PEMBULUH DARAH RETINA MENGGUNAKAN OTSU THRESHOLDING
Daftar Isi:
- Segmentation aims to separate the foreground and background. The proposed method consists of three stages, the preprocessing, segmentation, and post-processing. The preprocessing using Green Channel, Complement, contrast enhancement using Contrast Limited Adaptive Histogram Equalization and contrast adjustment. The segmentation using Otsu Thresholding. And the post-processing using Remove Small Object. The proposed method was tested on two publicly available datasets and reached an Accuracy 92.22%, Sensitivity 53.72%, Specificity 96,78% for STARE dataset and Accuracy 94.44%, Sensitivity 56.64%, Specificity 98,63% for DRIVE dataset.