File size: 4,201 Bytes
87782fd
 
e72950a
87782fd
e72950a
 
87782fd
 
 
 
e72950a
87782fd
 
e72950a
87782fd
e72950a
 
 
 
87782fd
e72950a
87782fd
e72950a
87782fd
e72950a
87782fd
e72950a
 
 
 
 
87782fd
e72950a
 
 
87782fd
e72950a
87782fd
e72950a
 
 
 
 
 
 
87782fd
 
e72950a
87782fd
e72950a
 
 
 
 
87782fd
 
e72950a
 
 
87782fd
e72950a
87782fd
e72950a
87782fd
e72950a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
87782fd
e72950a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
---
language:
  - zh
license: agpl-3.0
task_categories:
  - image-to-text
tags:
  - captcha
  - ocr
  - phpwind
pretty_name: PHPWind Captcha Dataset
---

# PHPWind Captcha Dataset

Labelled four-digit numeric captcha images for training and evaluating OCR on
authorized PHPWind deployments. This repository stores each visual captcha
family in an independent dataset directory so that samples from different
versions, forks, themes, or generators are never silently mixed.

**中文说明**: [README_zh.md](README_zh.md)

**Related model**: [FlanChanXwO/phpwind-captcha-ocr](https://huggingface.co/FlanChanXwO/phpwind-captcha-ocr)

## Dataset catalogue

| Dataset ID | Deployment or version identifier | Status | Images | Notes |
|---|---|---|---:|---|
| `legacy-footer-v0.7-beta` | Target deployment footer label `v0.7β` | Available | 997 | The only dataset currently published in this repository. The footer text is an observed deployment label, not a claim about an official PHPWind release version. |
| `phpwind-v7` | PHPWind v7 | Reserved | — | Create only after collecting and labelling an independent, authorized v7 sample set. |
| `phpwind-v8` | PHPWind v8 | Reserved | — | Create only after collecting and labelling an independent, authorized v8 sample set. |

`legacy-footer-v0.7-beta` is the training-data scope for the related model's
published reference metric. It does **not** establish compatibility with other
PHPWind releases, forks, themes, or captcha generators.

## Repository layout

```text
datasets/
├── index.json                         # Catalogue of published dataset directories
└── legacy-footer-v0.7-beta/
    ├── metadata.json                  # Source/version boundary and dataset properties
    ├── labels.json                    # Filename-to-label mapping
    └── images/                        # Captcha PNG files
```

Every future version-specific dataset must use the same isolated layout:

```text
datasets/phpwind-v7/
├── metadata.json
├── labels.json
└── images/
```

Do not place v7, v8, custom-theme, or fork samples in
`legacy-footer-v0.7-beta/images/`. Give each visually distinct generator its
own directory and catalogue entry.

## Current dataset details

| Property | Value |
|---|---|
| Dataset ID | `legacy-footer-v0.7-beta` |
| Captcha format | Four numeric digits (`0``9`) |
| Image count | 997 PNG files |
| Labels | One label per image in `labels.json` |
| Image path | `datasets/legacy-footer-v0.7-beta/images/` |
| Label path | `datasets/legacy-footer-v0.7-beta/labels.json` |
| Source boundary | Target deployment whose footer displayed `v0.7β` |

`labels.json` is an object keyed by filename. Each entry contains the final
four-digit `label`, the recorded `votes`, the number of votes (`count`), and
the agreement flag (`agreed`). Image paths are relative to the dataset
directory: `images/<filename>`.

```json
{
  "b1_0000.png": {
    "label": "9125",
    "votes": ["9125"],
    "count": 1,
    "agreed": true
  }
}
```

## Adding a new version-specific dataset

1. Create `datasets/<dataset-id>/images/`; use a distinct ID such as
   `phpwind-v7` or `phpwind-v8` only when that version has been established
   from authorized evidence.
2. Add only that generator's images and a separate `labels.json`. Keep the
   image filenames unique inside its own directory.
3. Add `metadata.json` with the observed version or deployment identifier,
   captcha format, image count, collection provenance, and any known visual
   customization. Do not infer an official version from a theme footer alone.
4. Add the dataset to `datasets/index.json` and update the catalogue table
   above. Record whether its labels have been independently reviewed.
5. Train and evaluate each version separately first. Combine datasets only in
   an explicitly documented multi-version experiment with per-version results.

## Responsible use and license

Use only captcha images from deployments you own or are explicitly authorized
to test. Do not use this dataset to automate account logins or bypass access
controls. The dataset is released under [GNU AGPL-3.0](https://www.gnu.org/licenses/agpl-3.0.html).