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.

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

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

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
RESULTS
Delivered an interactive segmentation pipeline that visualizes preprocessing stages and exports detected image regions for further inspection.
PROJECT LINKS
NOTE
The application performs image-region segmentation only. It does not perform OCR, handwriting recognition, language interpretation, or semantic text extraction.
