【paddleocr】发票ocr

发票文字识别解决方案：

使用paddleocr自带的模型进行文字检测+识别，根据发票内容信息特点提取所需的内容。比如需要提取下图中所标红框对应的内容（绿框中的信息）。

1 环境准备

paddleocr==2.4
Pillow==9.0.1

2 代码

from paddleocr import PaddleOCR, draw_ocr
from PIL import Image


def deal_dict(info_dict):
    if "number" in info_dict.keys():
        pass
    else:
        info_dict["number"] = ""
    if "date" in info_dict.keys():
        pass
    else:
        info_dict["date"] = ""
    if "rate" in info_dict.keys():
        pass
    else:
        info_dict["rate"] = ""
    if "amount" in info_dict.keys():
        pass
    else:
        info_dict["amount"] = ""
    if "tax" in info_dict.keys():
        pass
    else:
        info_dict["tax"] = ""
    i