AIM — EXAM OCR SYSTEM
Automated examination checking system developed for the AALAM school network (6 schools) in Kyrgyzstan to evaluate 20,000+ offline entrance examination sheets.
End-to-End 6-Step Automated Workflow
How AIM Exam OCR transforms physical paper tests into verified Excel results in seconds
Collect Sheets
Physical examination answer sheets are collected from entrance test halls across 6 AALAM schools.
High-Speed Scan
Batches of answer sheets are fed through high-speed sheetfed document scanners.
Image Transfer
Scanned images are transferred directly to the processing computer running the AIM software.
CV Processing
Hough alignment, homography rectification, and subpixel bubble snapping dynamically detect marked choices.
Auto Scoring
Answers are scored against the official answer keys, generating scores and candidate IDs automatically.
Excel Export
Final ranking, scores, and analytics are exported cleanly into a formatted Excel file for the school administration.
The Problem & The Automation Solution
AALAM schools conduct offline entrance examinations using physical paper answer sheets. More than 20,000 examination sheets needed to be evaluated. Checking such a vast volume manually would demand countless hours of tedious teacher labor with inevitable grading mistakes.
The Problem
20,000+ paper answer sheets across 6 schools. Manual grading caused severe delays, teacher burnout, and error susceptibility.
The Engineering Solution
Our 2-person team developed a custom desktop CV pipeline that digests scanner feeds, rectifies sheet tilt via homography, dynamically snaps to bubbles, and exports to Excel in real-time.
