How to Scan OMR Sheets
Scan filled OMR sheets with any standard scanner for automated grading using our Dynamic OCR Algorithm
How to Scan OMR Sheets
After students complete their paper-based exams, scan the filled OMR sheets to automatically grade them. Course37's Dynamic OCR Algorithm achieves 99%+ accuracy across all paper sets.
Prerequisites
- A standard flatbed or sheet-feed scanner (any brand works)
- Scanning at 300 DPI recommended (minimum 200 DPI)
- Grayscale or Color mode (both work; grayscale is faster)
- Filled OMR sheets from your OMR template
Step-by-Step Scanning Process
Step 1 — Scan the Sheets
- Collect all filled OMR sheets from the exam
- Scan them using your scanner as:
- Individual images — JPEG or PNG (one file per sheet)
- Multi-page PDF — All sheets in one PDF file
- Recommended: 300 DPI, grayscale
Step 2 — Upload to Course37
- Go to Exams → OMR → Scan Results
- Select the Exam this scan belongs to
- Select the OMR Template that was used
- Upload the scanned files — drag and drop supported
- You can upload multiple files at once
📸 Screenshot: OMR upload interface with drag-and-drop
Step 3 — Automated Processing
Our Dynamic OCR Algorithm processes each sheet:
- Detects orientation — Handles slightly rotated or misaligned papers
- Identifies student — Reads the student ID/roll number field
- Detects filled bubbles — Reads each question's selected answer
- Identifies set — Determines which set (A, B, C, D) was used
- Calculates score — Compares answers against the answer key
📸 Screenshot: Processing progress with detected answers overlay
Step 4 — Review Results
- After processing, review the parsed results
- Each sheet shows:
- Scanned image overlay with detected marks
- Detected vs. expected answers per question
- Confidence level for each detection
- Flag low-confidence detections — These need manual verification
- Correct any misdetections manually if needed
Step 5 — Confirm
- After reviewing, click Confirm Results
- Results are saved and ready for publishing
Accuracy & Edge Cases
The Dynamic OCR Algorithm handles challenging scenarios:
| Scenario | How It's Handled |
|---|---|
| Partially filled bubbles | Detected based on fill density analysis |
| Multiple bubbles selected | Flagged for manual review |
| Erasure marks | Differentiated from actual selections |
| Misaligned papers | Auto-corrected using template markers |
| Different sets (A/B/C/D) | Automatically detected and graded against correct answer key |
| Low quality scan | Processing still works; lower confidence items flagged |
Overall accuracy: 99%+ across all paper sets and scanning conditions.
Scanning Tips for Best Results
- Use 300 DPI — Higher resolution = higher accuracy
- Clean scanner glass — Dust and smudges affect detection
- Flatten sheets — Crumpled papers may scan poorly
- Consistent orientation — Feed all sheets the same way
- Batch scan — Use a sheet-feed scanner for faster processing
Related Articles
- How to Create an OMR Template — Design answer sheets
- How to Manage OMR Results — Review and publish results
- How to See Results — Exam results management
OMRscanscannerOCRgradingautomatedaccuracy