---
id: qrdet
version: "2.5"
license: MIT
license_treatment: permissive
maintenance: aging
---

# qrdet — Robust QR Detector based on YOLOv8

License: permissive · Maintenance: aging · Downloads: 81.6K/mo

## What it is and what it does

QRDet is a QR code detector built on YOLOv8 that finds and segments QR codes in images, even when they are rotated, partially obscured, or in challenging lighting. It returns detection results as dictionaries containing bounding boxes, confidence scores, and precise polygon coordinates that outline the QR code's shape. The package offers four model sizes (n, s, m, l) to trade off speed against detection capability, and allows tuning of confidence thresholds and non-maximum suppression parameters to control false positives and duplicate detections.

The core dependency chain includes ultralytics (for YOLOv8), numpy (for array operations), quadrilateral-fitter (for polygon refinement), requests (for model weight downloads), and tqdm (for progress feedback). It accepts images as numpy arrays, PIL Images, torch Tensors, file paths, URLs, or screenshots. The package is designed for detection only; if you need both detection and decoding, the description suggests looking at QReader instead.

Use it for:

- Locating QR codes in product photos or inventory images for automated scanning workflows.
- Extracting precise QR code boundaries in rotated or skewed images before passing to a decoder.
- Building a QR detection pipeline in computer vision applications where confidence scores and segmentation polygons are needed.
- Processing batches of images to find all QR codes and their exact positions for downstream processing.
- Detecting QR codes in challenging conditions where simpler detectors fail.

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Detects and segments QR codes in images using a YOLOv8-based model, returning bounding boxes, confidence scores, and precise polygon coordinates even in difficult positions.

Yes, if you need robust QR detection with polygon-level precision. The package is stable (Production/Stable classifier), permissively licensed, has no known vulnerabilities, and offers flexible model sizing. The aging maintenance status (last release mid-2024) is a minor concern but not a blocker if the current version meets your needs. Install friction is low. Not suitable if you also need QR decoding in the same package.

## Install

pip install qrdet
uv add qrdet
poetry add qrdet

## Installing qrdet

Before you install: Low friction install with a pure-Python wheel. Depends on ultralytics, numpy, requests, tqdm, and quadrilateral-fitter. Maintenance is aging—last release was 2024-06-15 and the repo has not been updated since 2025-02-23, though it remains active and unarchived.

License in practice: MIT license (permissive) places no restrictions on use, modification, or distribution in commercial or private projects.

Quickstart:

pip install qrdet

from qrdet import QRDetector

detector = QRDetector(model_size='s')
detections = detector.detect(image='path/to/image.jpg')
for det in detections:
    print(det['bbox_xyxy'], det['confidence'])

Model weights are downloaded on first use; ensure the weights_folder (default: package/.model) is writable, or specify an alternative writable path (e.g., /tmp on AWS Lambda).

Verify before relying:

- Whether model weights are downloaded on first use and their total size impact on installation footprint.
- Performance characteristics (inference speed, accuracy) across the four model sizes ('n', 's', 'm', 'l').
- Compatibility with different Python versions—requires_python is unspecified in the fact sheet.

## Package facts

- License: MIT (permissive)
- Python support: unspecified
- Install friction: low
- Maintenance: aging
- Downloads: 81.6K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

qr code detection, qr code localization, yolov8 qr detector, qr segmentation, detect qr codes in images, qr bounding box, qr polygon detection, qr-code, object-detection, computer-vision

[View on SkillFed](https://skillfed.io/packages/qrdet) · [View on PyPI](https://pypi.org/project/qrdet/)