Skip to content
#

otsu-thresholding

Here are 70 public repositories matching this topic...

A lightweight, heuristic-based algorithm for segmenting characters in Iranian license plates using OpenCV. Features robust handling of shadows, noise, and connected characters without Deep Learning, achieving 98.68% accuracy.

  • Updated Dec 16, 2025
  • Python

This DR detection methodology has six steps: preprocessing, segmentation of blood vessels, segmentation of OD, detection of MAs and hemorrhages, feature extraction and classification. For segmentation of blood vessels BCDU-Net is used. For OD segmentation, U-Net model is used. MAs and hemorrhages are extracted using Otsu thresholding technique. …

  • Updated Oct 2, 2020
  • Jupyter Notebook

Add this topic to your repo

To associate your repository with the otsu-thresholding topic, visit your repo's landing page and select "manage topics."

Learn more