~shahper.shahin
Back to Academic Showcase
Technical ReportOptical Systems & Artificial Intelligence • Patent Application #202511052205July 2025

System for Non-Destructive Highlighting & Processing of Physical Text

Abstract

“Formal patent specification for a non-destructive digital highlighter integrating optical text capture, color-space segmentation, optical character recognition, and AI-based contextual note organization.”

PythonOpenCVTesseract OCRPatent LawImage Processing

Research / Coursework Objective

Enable physical document interaction without ink damage while synchronizing structured notes into digital knowledge graphs.

Methodology & Experimental Procedure

  • 1.Optical hardware calibration and document alignment
  • 2.Adaptive HSV color segmentation for multi-tint highlight detection
  • 3.Bounding box spatial clustering for multiline paragraph extraction
  • 4.Cloud synchronization and LLM contextual indexing

Technical Implementation

Built using Python, OpenCV, and Tesseract OCR with RESTful sync endpoints.

Key Findings & Analytical Results

  • Color-coded highlighting maps cleanly to semantic categories in digital knowledge graphs
  • Adaptive thresholding achieves 98.4% accuracy under varying ambient lighting
Academic References
  • [1] Indian Patent Application No. 202511052205 (Filed July 2025)
  • [2] Digital Image Processing (Gonzalez & Woods)