Ashwin Mishra
AI/ML developer · New Delhi, India

Evidence Analysis: telling site photos from documents

February 6, 2026 · #computer-vision #ocr #iit-kanpur

Evidence-submission portals often receive the wrong kind of upload. People submit a document where a site photo belongs, or the other way round, and catching these by hand doesn't scale.

The pipeline scores each image on five signals: pixel variance, edge density, entropy, heuristic rules and OCR layout analysis. Four document indicators each add a point: text-area ratio ≥ 0.12, vertical text spread ≥ 0.4, at least 4 valid words, and consistent text alignment. A voting step using 40% region agreement and OCR feasibility resolves ambiguous cases.

Stack: Python, FastAPI, React, OCR.

Code: github.com/Ashwin07Mishra/Evidence-Analysis

← back to case studies


This is the plain-HTML version of this site — no JavaScript, CSS 2.1 only, built to render in Dillo.
Full version of this page · github · linkedin · resume.pdf · ashwin.mishra07@gmail.com