Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.qmr.kr/llms.txt

Use this file to discover all available pages before exploring further.

Scanning is the step that turns a stack of completed paper answer sheets into graded results in your dashboard. QMR is designed so that the only thing you need to do is provide the image — everything else (identifying the exam, finding the student, reading the answers, calculating the score) happens automatically. This page explains how that works, what upload options you have, and what to expect if something goes wrong.

​
How QR codes identify everything automatically

Every answer sheet printed by QMR contains a unique QR code. When you upload a sheet, QMR reads this code before it does anything else. The QR code encodes three things:
  • Which exam the sheet belongs to
  • Which round (class group) it was printed for
  • Which student should be credited with the result
Because of this, there is no step where you select the exam or enter the student’s name before scanning. You simply upload the image and QMR routes the result to the right place automatically. This also means that sheets from different exams or different rounds can be uploaded together — QMR sorts them correctly.
If the QR code on a sheet cannot be read (for example, because it is torn, heavily creased, or obscured by a stray mark), the scan will fail and no submission will be created for that sheet. QMR will flag the failed scan so you can retry with a cleaner image or re-scan the sheet.

​
AI recognition

QMR’s image recognition is built to handle real classroom conditions, not studio-quality scans. The AI corrects for:
  • Shadows and uneven lighting — from overhead lights, windows, or a hand holding the camera.
  • Skew and perspective distortion — when a sheet is photographed at an angle rather than straight-on.
  • Wrinkles and folds — from sheets being carried in bags or folded for collection.
  • Ambiguous marks — when a student’s mark is faint, partially erased, or between two bubbles, QMR makes a best-effort determination based on the relative density of the mark.
QMR claims a 99.9%+ recognition rate across real-world answer sheets. For the rare sheet that produces an unexpected result, you can reprocess it (see below) or manually adjust the submission from the submissions tab.

​
Three ways to upload answer sheets

Open the Scan section of the dashboard to choose your upload method.
On a mobile device, a Take photo button appears alongside the file upload options. Tap it to open your device camera pointing at the answer sheet. QMR uploads the image immediately after capture and begins processing.This is the fastest option when you are collecting sheets in the classroom — you can scan each sheet one by one as students hand them in.

​
What happens after you upload

The full processing flow takes a few seconds per sheet:
1

Upload

Your file is transferred to QMR’s cloud storage. A progress indicator appears while the upload is in progress.
2

QR code read

QMR locates and decodes the QR code on the sheet to identify the exam, the round, and the student.
3

Answer extraction

The AI reads each problem marking area on the sheet and records the answer the student marked.
4

Submission created

A graded submission is added to the exam’s submissions list. The student’s score is calculated against the answer key immediately if the key is already complete. If you have not yet entered the answer key, the submission is stored and scored automatically once you do.
For a PDF batch upload, a progress panel tracks how many pages have been processed and how many remain, so you can monitor a large batch in real time.

​
Reprocessing a scan

If a scan fails — for example because the QR code was unreadable or you uploaded the wrong image — you can reprocess it. Go to the Scans page in your teacher dashboard, find the failed scan request, and upload a cleaner image of the sheet to create a new scan request. The failed entry will be updated once the new attempt completes.
For the best recognition results, photograph sheets on a flat, well-lit surface with the camera held directly above the sheet. Avoid casting shadows with your hand or phone case.