Metadata-Version: 2.1
Name: mtgscan
Version: 1.0.8
Summary: Convert an image containing Magic cards to decklist
Home-page: https://github.com/fortierq/MTGScan
License: MIT
Keywords: mtg,scan,OCR,magic,cards,computer vision
Author: qfortier
Author-email: qpfortier@gmail.com
Requires-Python: >=3.8,<4.0
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.8
Classifier: Programming Language :: Python :: 3.9
Requires-Dist: matplotlib (>=3.3.3,<4.0.0)
Requires-Dist: numpy (>=1.23.0,<2.0.0)
Requires-Dist: psutil (>=5.8.0,<6.0.0)
Requires-Dist: requests (>=2.25.0,<3.0.0)
Requires-Dist: symspellpy (>=6.7.0,<7.0.0)
Project-URL: Repository, https://github.com/fortierq/MTGScan
Description-Content-Type: text/markdown

# MTGScan
[![License: MIT](https://img.shields.io/badge/License-MIT-green.svg)](https://opensource.org/licenses/MIT)
[![CodeFactor](https://www.codefactor.io/repository/github/fortierq/mtgscan/badge)](https://www.codefactor.io/repository/github/fortierq/mtgscan)

![mtgscan](https://user-images.githubusercontent.com/49362475/102022934-448ffb80-3d8a-11eb-8948-3a10d190162a.jpg)

MTGScan uses OCR recognition to list Magic cards from an image.  
After OCR, cards are looked up in a dictionnary provided by [MTGJSON](https://mtgjson.com), using fuzzy search with [SymSpell](https://github.com/wolfgarbe/SymSpell).

## [Try the Web App](https://qfmtgscanapp.azurewebsites.net)
```mermaid
  flowchart LR;
  subgraph "Browser"
    C1[Client];
    C2[Client];
    C3[Client];
  end
  subgraph "Frontend"
    F((Flask Server));
  end
  subgraph "Backend"
    W1[Celery Worker<br>using mtgscan];
    W2[Celery Worker<br>using mtgscan];
  end
  subgraph "Cloud"
    A[Azure Read OCR];
  end
  C1 <-->|Socket.IO| F;
  C2 <-->|Socket.IO| F;
  C3 <-->|Socket.IO| F;
  F <-->|RedisMQ| W1;
  F <-->|RedisMQ| W2;
  W1 <-->|API| A;
  W2 <-->|API| A;
```
[Repository for the web app](https://github.com/fortierq/mtgscan-app)

## Prerequisites

- Python >= 3.7
- Credentials for the required OCR (e.g Azure Computer Vision Read API)

## Installation

### ... with Poetry

```python
poetry install
```

### ... with requirements

```python
pip install -r requirements.txt
```

### ...  with pip

```console
pip install mtgscan
```

## OCR

Currently, only Azure OCR is supported. To add an OCR, inherit mtgscan.ocr.OCR.  

### Azure

API subscription key and endpoint must be stored in environment variables `AZURE_VISION_KEY` and `AZURE_VISION_ENDPOINT` respectively.  
Steps:
- Subscribre for a free Azure account: https://azure.microsoft.com/free/cognitive-services
- Create a Computer Vision resource: https://portal.azure.com/#create/Microsoft.CognitiveServicesComputerVision
- Get your key and endpoint

## Tests

Every test case is stored in a separated folder in tests/samples/ containing:
- image.*: image of Magic cards
- deck.txt: decklist of the cards on the image

To run every test:
```python
poetry run python tests/test.py
```

This produces the following outputs, for each sample and OCR:
- statistics about number of cards found, number of errors...
- test.log: informations about the run
- errors.txt: history of the number of errors made by the OCR
- box_texts.txt: output of the OCR

## Basic usage

Let's retrieve the decklist from the following screenshot:
![Screenshot](https://user-images.githubusercontent.com/49362475/105632710-fa07a180-5e54-11eb-91bb-c4710ef8168f.jpeg)

```python
from mtgscan.text import MagicRecognition
from mtgscan.ocr.azure import Azure

azure = Azure()
rec = MagicRecognition(file_all_cards="all_cards.txt", file_keywords="Keywords.json")  # download card files from mtgjson if missing
box_texts = azure.image_to_box_texts("https://user-images.githubusercontent.com/49362475/105632710-fa07a180-5e54-11eb-91bb-c4710ef8168f.jpeg")
deck = rec.box_texts_to_deck(box_texts)
print(deck)
```

Output:
```console
4 Ancient Tomb
4 Mishra's Factory
4 Mishra's Workshop
1 Strip Mine
1 Tolarian Academy
4 Wasteland
1 Sacrifice
1 Mox Ruby
1 Mox Emerald
1 Mox Jet
1 Mox Pearl
1 Mox Sapphire
1 Black Lotus
1 Mana Crypt
1 Sol Ring
4 Phyrexian Revoker
4 Arcbound Ravager
1 Thorn of Amethyst
4 Sphere of Resistance
4 Foundry Inspector
3 Chief of the Foundry
1 Trinisphere
1 Lodestone Golem
1 Mystic Forge
2 Fleetwheel Cruiser
1 Traxos, Scourge of Kroog
4 Walking Ballista
3 Stonecoil Serpent
1 Chalice of the Void

3 Mindbreak Trap
4 Leyline of the Void
2 Crucible of Worlds
4 Pithing Needle
2 Wurmcoil Engine
```

## Task list
- [x] Tested on MTGO, Arena and IRL (simple) images
- [x] Handle sideboard (only on the right side)  
- [x] Support for stacked cards
- [ ] Add and compare OCR (GCP, AWS...)

