| .. |
|
dict2md
|
473a0a7de0
拼接markdown时,如果para_text为空则跳过拼接
|
1 year ago |
|
filter
|
d5dbed7325
目录重构
|
1 year ago |
|
layout
|
d5dbed7325
目录重构
|
1 year ago |
|
libs
|
f36c26565e
使用面积占比方式判断一行文本是不是在一个layoutbox里
|
1 year ago |
|
para
|
d3c9cb84f8
分段部分log限定在debug模式下才能输出
|
1 year ago |
|
post_proc
|
71a042d9fc
footnote检测逻辑更新
|
1 year ago |
|
pre_proc
|
90ea9096e5
部分layout最底部的文本圈不全,降低阈值减少底边文本被丢弃的可能性
|
1 year ago |
|
train_utils
|
05161c6e62
feat: backup footnote_bbox_tmp
|
1 year ago |
|
__init__.py
|
d5dbed7325
目录重构
|
1 year ago |
|
pdf_parse_by_model.py
|
f68c66290c
update code
|
1 year ago |
|
pdf_parse_by_ocr.py
|
3d2fcc9dce
删除无用代码
|
1 year ago |
|
pdf_parse_for_train.py
|
05161c6e62
feat: backup footnote_bbox_tmp
|
1 year ago |
|
pipeline.py
|
07e4f115e6
ocr_pdf_intermediate_dict_to_markdown_with_para输出nlp格式的markdown
|
1 year ago |