PROJECTS

License Plate Character Detection

Industrial Machine Vision

Overview

Developed a computer vision pipeline for detecting and segmenting characters on vehicle license plates. The project emphasizes classical image processing techniques for robust feature extraction and segmentation.

Technical Features

Image Preprocessing
  • Applied grayscale conversion, noise reduction, and normalization to prepare images for analysis.
  • Used thresholding and adaptive binarization for robust character segmentation under varying lighting conditions.
Feature Extraction & Segmentation
  • Implemented morphological operations (erosion, dilation, opening, closing) to enhance character structures.
  • Applied contour analysis to detect and isolate individual characters.
  • Filtered candidate regions using size, aspect ratio, and geometric constraints for accurate detection.
Pipeline Integration
  • Developed a modular pipeline combining preprocessing, segmentation, and filtering for end-to-end detection.
  • Designed for batch processing and scalability to handle multiple license plate images efficiently.

Tech Stack

Have a Challenge? Let’s Solve It.

Copyright © 2025 Saurabh Singh