营业执照信息识别

标签：info code 营业执照 信息 replace strip credit ocr 识别

python依赖

paddlepaddle==2.3.1
paddleocr==2.5.0.3

python示例

#!/user/bin/env python
# coding=utf-8


from paddleocr import PaddleOCR


class BusinessLicense:

    def __init__(self, img, **kwargs):
        self.ocr_cls = PaddleOCR(use_angle_cls=kwargs.get("use_angle_cls", True), use_gpu=kwargs.get("use_gpu", False),
                                 lang=kwargs.get("use_gpu", "ch"))
        self.img = img
        self.flag = False

    def verification_credit_code(self, credit_code):
        if credit_code:
            _sum = 0
            code_dict = {
                '0': 0, '1': 1, '2': 2, '3': 3, '4': 4, '5': 5, '6': 6, '7': 7, '8': 8, '9': 9,
                'A': 10, 'B': 11, 'C': 12, 'D': 13, 'E': 14, 'F': 15, 'G': 16, 'H': 17, 'J': 18, 'K': 19, 'L': 20,
                'M': 21,
                'N': 22, 'P': 23, 'Q': 24,
                'R': 25, 'T': 26, 'U': 27, 'W': 28, 'X': 29, 'Y': 30}

            weights = ['1', '3', '9', '27', '19', '26', '16', '17', '20', '29', '25', '13', '8', '24', '10', '30', '28']

            for i in range(len(credit_code) - 1):
                _sum += code_dict[credit_code[i]] * int(weights[i])

            mod = 31 - _sum % 31

            if (mod == code_dict[credit_code[-1]]) or ((mod == 31) and code_dict[credit_code[-1]] == 0):
                self.flag = True
        return self.flag

    def parse(self):
        credit_code = ""
        company_name = ""
        address = ""
        legal_person = ""
        ocr_cls = PaddleOCR(use_angle_cls=True, use_gpu=False, lang="ch")
        ocr_result = ocr_cls.ocr(self.img, cls=True)
        for i, x in enumerate(ocr_result):
            info = x[-1][0]
            if info.startswith("统一社会信用代码"):
                credit_code = info.split("统一社会信用代码")[1].strip()
            elif info.find("公司") != -1:
                if info.startswith("称") and info.endswith("公司"):
                    company_name = info.replace(":", "").replace(":", "")[1:].strip()
                elif info.startswith("名称") and info.endswith("公司"):
                    company_name = info.replace(":", "").replace(":", "")[2:].strip()
                elif info.endswith("公司"):
                    company_name = info.strip()
            elif info.startswith("住"):
                if info == "住" and i < len(ocr_result) - 1:
                    if ocr_result[i + 1][-1][0].replace(":", "").replace(":", "") == "所":
                        address = ocr_result[i + 2][-1][0].replace(":", "").replace(":", "").strip()
                elif info == "住所" and i < len(ocr_result) - 1:
                    address = ocr_result[i + 1][-1][0].replace(":", "").replace(":", "").strip()
                elif info.startswith("住所"):
                    address = info.split("住所").replace(":", "").replace(":", "").strip()
            elif info.startswith("法定代表人"):
                if info == "法定代表人":
                    legal_person = ocr_result[i + 1][-1][0].replace(":", "").replace(":", "").strip()
                else:
                    legal_person = info.replace(":", "").replace(":", "").split("法定代表人")[1].strip()

        return {"company_name": company_name,
                "credit_code": credit_code,
                "address": address,
                "legal_person": legal_person}


img_path = "./zhi.jpeg"
bl = BusinessLicense(img=img_path)
data = bl.parse()
print(data)
print(bl.verification_credit_code(data.get("unified_social_credit_code")))

　　

标签：info,code,营业执照,信息,replace,strip,credit,ocr,识别
From： https://www.cnblogs.com/navysummer/p/16583898.html