CDLA是一个中文文档版面分析数据集,面向中文文献类(论文)场景。包含以下10个label:
正文 | 标题 | 图片 | 图片标题 | 表格 | 表格标题 | 页眉 | 页脚 | 注释 | 公式 |
---|---|---|---|---|---|---|---|---|---|
Text | Title | Figure | Figure caption | Table | Table caption | Header | Footer | Reference | Equation |
共包含5000张训练集和1000张验证集,分别在train和val目录下。每张图片对应一个同名的标注文件(.json)。
样例展示:
-
百度云下载:https://pan.baidu.com/s/1449mhds2ze5JLk-88yKVAA, 提取码: tp0d
-
Google Drive Download:https://drive.google.com/file/d/14SUsp_TG8OPdK0VthRXBcAbYzIBjSNLm/view?usp=sharing
我们的标注工具是labelme,所以标注格式和labelme格式一致。这里说明一下比较重要的字段。
"shapes": shapes字段是一个list,里面有多个dict,每个dict代表一个标注实例。
"labels": 类别。
"points": 实例标注。因为我们的标注是Polygon形式,所以points里的坐标数量可能大于4。
"shape_type": "polygon"
"imagePath": 图片路径/名
"imageHeight": 高
"imageWidth": 宽
展示一个完整的标注样例:
{
"version":"4.5.6",
"flags":{},
"shapes":[
{
"label":"Title",
"points":[
[
553.1111111111111,
166.59259259259258
],
[
553.1111111111111,
198.59259259259258
],
[
686.1111111111111,
198.59259259259258
],
[
686.1111111111111,
166.59259259259258
]
],
"group_id":null,
"shape_type":"polygon",
"flags":{}
},
{
"label":"Text",
"points":[
[
250.5925925925925,
298.0740740740741
],
[
250.5925925925925,
345.0740740740741
],
[
188.5925925925925,
345.0740740740741
],
[
188.5925925925925,
410.0740740740741
],
[
188.5925925925925,
456.0740740740741
],
[
324.5925925925925,
456.0740740740741
],
[
324.5925925925925,
410.0740740740741
],
[
1051.5925925925926,
410.0740740740741
],
[
1051.5925925925926,
345.0740740740741
],
[
1052.5925925925926,
345.0740740740741
],
[
1052.5925925925926,
298.0740740740741
]
],
"group_id":null,
"shape_type":"polygon",
"flags":{}
},
{
"label":"Footer",
"points":[
[
1033.7407407407406,
1634.5185185185185
],
[
1033.7407407407406,
1646.5185185185185
],
[
1052.7407407407406,
1646.5185185185185
],
[
1052.7407407407406,
1634.5185185185185
]
],
"group_id":null,
"shape_type":"polygon",
"flags":{}
}
],
"imagePath":"val_0031.jpg",
"imageData":null,
"imageHeight":1754,
"imageWidth":1240
}
执行命令:
# train
python3 labelme2coco.py CDLA_dir/train train_save_path --labels labels.txt
# val
python3 labelme2coco.py CDLA_dir/val val_save_path --labels labels.txt
转换结果保存在train_save_path/val_save_path目录下。
labelme2coco.py取自labelme,更多信息请参考labelme官方项目