Langchain-Chatchat/loader/image_loader.py
zhenkaivip dd93837343
使用paddleocr实现 (#342)
* jpg and png ocr

* fix

* write docs to tmp file

* fix

* [BUGFIX] local_doc_qa.py line 172: logging have no end args. (#323)

* image loader

* fix

* fix

* update api.py

* update api.py

* update api.py

* update README.md

* update api.py

* add pdf_loader

* fix

---------

Co-authored-by: RainGather <3255329+RainGather@users.noreply.github.com>
Co-authored-by: imClumsyPanda <littlepanda0716@gmail.com>
2023-05-13 08:45:17 +08:00

29 lines
1.2 KiB
Python

"""Loader that loads image files."""
from typing import List
from langchain.document_loaders.unstructured import UnstructuredFileLoader
from paddleocr import PaddleOCR
import os
class UnstructuredPaddleImageLoader(UnstructuredFileLoader):
"""Loader that uses unstructured to load image files, such as PNGs and JPGs."""
def _get_elements(self) -> List:
def image_ocr_txt(filepath, dir_path="tmp_files"):
if not os.path.exists(dir_path):
os.makedirs(dir_path)
filename = os.path.split(filepath)[-1]
ocr = PaddleOCR(lang="ch", use_gpu=False, show_log=False)
result = ocr.ocr(img=filepath)
ocr_result = [i[1][0] for line in result for i in line]
txt_file_path = os.path.join(dir_path, "%s.txt" % (filename))
with open(txt_file_path, 'w', encoding='utf-8') as fout:
fout.write("\n".join(ocr_result))
return txt_file_path
txt_file_path = image_ocr_txt(self.file_path)
from unstructured.partition.text import partition_text
return partition_text(filename=txt_file_path, **self.unstructured_kwargs)