Upload 114 files

README.md

@@ -1,3 +1,32 @@
----
-license: mit
----
+<p align="center">
+  <a href="README.md">English</a> | <a href="README_ZH.md">中文</a>
+</p>
+
+# Introduction
+The original images of this dataset are from [DocLayNet-v1.2](https://huggingface.co/datasets/docling-project/DocLayNet-v1.2). You can run `src/1_save_img_from_parquet.py` to convert the parquet files into images.
+
+# Page Number Cropping
+This dataset uses the `yolo11-medium-page-best-quant-ir9.onnx` weights from [Here](https://huggingface.co/wqzh/DocLayout-PageNumDet/tree/main/models) to detect page number bounding boxes from the images obtained above, and then crop out the page number regions (detection errors may exist, as no post-processing was applied by the author).
+
+The cropped images are saved in the `images-crops` directory, and the detection results are saved in the `images-labels` directory. The author added a constraint: the YOLO-format label file is only saved when exactly one page number box is detected, while crops of all detected boxes are retained. Therefore, the number of files in `images-crops` is greater than that in `images-labels`.
+
+You can run `src/2_det_onnx_wo_opencv_cuda.py` to detect page numbers, crop images, and save label files.
+
+```shell
+find images-labels -type f -name "*.txt" | wc -l
+find images-crops -type f -name "*.jpg" | wc -l
+```
+- Number of `.txt` files in `images-labels`: 15902
+- Number of `.jpg` files in `images-crops`: 17203
+
+# Page Number Concatenation
+`src/3_concat_crops_by_pages.py` concatenates all cropped images from `images-crops` and saves them in the `images-crops-concat` directory, so you can preview the entire dataset.
+
+Example
+<img src="images-crops-concat/train-00005_p1.png" width="500">
+
+Note: If you intend to use the label files from this dataset, they need to be cleaned separately.
+
+# Page Number Coordinate Distribution
+`src/4_visualize_yolo_txt_distribution.py` visualizes the distribution of bounding box centers from all label files in `images-labels`.
+<img src="yolo_centers_distribution.png" width="500">

README_ZH.md

@@ -0,0 +1,34 @@
+<p align="center">
+  <a href="README_ZH.md">中文</a> | <a href="README.md">English</a>
+</p>
+
+# 简介
+本数据集的原始图片来源于[DocLayNet-v1.2](https://huggingface.co/datasets/docling-project/DocLayNet-v1.2)。 可以运行`src/1_save_img_from_parquet.py` 把parquet文件转换成图片。
+
+# 裁剪页码
+本数据集使用[Here](https://huggingface.co/wqzh/DocLayout-PageNumDet/tree/main/models)的`yolo11-medium-page-best-quant-ir9.onnx`权重，从上一步得到的图像中提取出页码的坐标，并裁剪出页码图片（检测可能存在错误，作者没有做任何后处理）。
+
+裁剪图片保存在`images-crops`目录下,检测的结果保存在`images-labels`目录下。作者做了一个限制：只有检测的页码框数量为1的时候，才会保存txt格式的标注文件，而检测的所有页码框的截图会全部保留。因此，`images-crops`目录下的文件数量 会大于 `images-labels`目录下的文件数量。
+
+可以运行`src/2_det_onnx_wo_opencv_cuda.py` 检测图片,裁剪图片，保存标注文件。
+
+```shell
+find images-labels -type f -name "*.txt" | wc -l
+find images-crops -type f -name "*.jpg" | wc -l
+```
+- images-labels 目录下 .txt 文件数: 15902
+- images-crops 目录下 .jpg 文件数: 17203
+
+# 页码合并
+`src/3_concat_crops_by_pages.py`将`images-crops`下所有的图片进行合并，并保存在`images-crops-concat`目录下。因此你可以预览整个数据集。
+
+示例
+<img src="images-crops-concat/train-00005_p1.png" width="500">
+
+注意：如果要使用这个数据集的label文件，需要单独清洗。
+
+# 页码坐标分布
+`src/4_visualize_yolo_txt_distribution.py`将`images-labels`下所有的标注文件的坐标分布可视化。
+<img src="yolo_centers_distribution.png" width="500">
+
+