Skip to content
ALL PROJECTS
COMPUTER VISION // IMAGE SEGMENTATION // 11

Word Segmentation OpenCV.

Built a Streamlit and OpenCV application that segments word-like image regions through grayscale conversion, inverse thresholding, dilation, contour detection, and bounding-box extraction.

CLIENT

Academic CV Project

ROLE

Computer Vision Developer

CATEGORY

COMPUTER VISION // IMAGE SEGMENTATION

STATUS

Live

LIVE DEMO

GITHUB

Word Segmentation OpenCV cover preview

CHALLENGE

Document images require preprocessing and region grouping before individual word-like components can be isolated for inspection.

Word Segmentation OpenCV mockup 1

SOLUTION

Implemented adjustable preprocessing, line and word dilation, contour sorting, bounding-box visualization, cropped-region output, and downloadable processed images.

Word Segmentation OpenCV mockup 2

KEY FEATURES

  • 01Image upload
  • 02Image resizing
  • 03Grayscale conversion
  • 04Inverse thresholding
  • 05Line-region dilation
  • 06Word-region dilation
  • 07Contour detection and sorting
  • 08Bounding-box extraction
  • 09Segmented-region output
  • 10Processed-image download

TECHNOLOGY

PythonOpenCVNumPyPillow

RESULTS

Delivered an interactive segmentation pipeline that visualizes preprocessing stages and exports detected image regions for further inspection.

PROJECT LINKS

LIVE DEMOVISIT
GITHUBVIEW

NOTE

The application performs image-region segmentation only. It does not perform OCR, handwriting recognition, language interpretation, or semantic text extraction.

PROJECT SCOPEREADY FORTECHNICAL REVIEW.