Skip to content

周伟奇 / bmw-ocr

Go to a project
Toggle navigation
Toggle navigation pinning
7f868e1c by 周伟奇
Options

contract cdfl

1 parent 4d0b881c
Inline Side-by-side
Showing 2 changed files with 62 additions and 5 deletions
src/common/electronic_afc_contract/get_char.py
......@@ -774,6 +774,9 @@ class Finder:
# 这是为了同时兼容 8.1 版本
if borrower_name['words'] == None:
borrower_name, borrower_id = self.get_somebody(top='借款人及抵押人：', bottom='共同借款人及共同抵押人：')
# 这是为了兼容车贷分离版本
if borrower_name['words'] == None:
borrower_name, borrower_id = self.get_somebody(top='借款人：', bottom='共同借款人及抵押人：')
self.init_result['page_2']['借款人及抵押人']['name'] = borrower_name
self.init_result['page_2']['借款人及抵押人']['id'] = borrower_id
# 找共同借款人及共同抵押人
......@@ -852,10 +855,16 @@ class Finder:
self.init_result['page_7']['合同编号'] = contract_no
signature_name, signature_date = self.get_last_page_signature(page_num='6',
top='合同编号', bottom='共同借款人')
if signature_name['words'] == None:
signature_name, signature_date = self.get_last_page_signature(page_num='6',
top='合同编号', bottom='共同借款人（抵押人）')
self.init_result['page_7']['主借人签字']['签字'] = signature_name
self.init_result['page_7']['主借人签字']['日期'] = signature_date
signature_name, signature_date = self.get_last_page_signature(page_num='6',
top='共同借款人', bottom='保证人1')
if signature_name['words'] == None:
signature_name, signature_date = self.get_last_page_signature(page_num='6',
top='共同借款人（抵押人）', bottom='保证人1')
self.init_result['page_7']['共借人签字']['签字'] = signature_name
self.init_result['page_7']['共借人签字']['日期'] = signature_date
signature_name, signature_date = self.get_last_page_signature(page_num='6',
......@@ -881,10 +890,16 @@ class Finder:
self.init_result['page_8']['合同编号'] = contract_no
signature_name, signature_date = self.get_last_page_signature(page_num='7',
top='合同编号', bottom='共同借款人')
if signature_name['words'] == None:
signature_name, signature_date = self.get_last_page_signature(page_num='7',
top='合同编号', bottom='共同借款人（抵押人）')
self.init_result['page_8']['主借人签字']['签字'] = signature_name
self.init_result['page_8']['主借人签字']['日期'] = signature_date
signature_name, signature_date = self.get_last_page_signature(page_num='7',
top='共同借款人', bottom='保证人1')
if signature_name['words'] == None:
signature_name, signature_date = self.get_last_page_signature(page_num='7',
top='共同借款人（抵押人）', bottom='保证人1')
self.init_result['page_8']['共借人签字']['签字'] = signature_name
self.init_result['page_8']['共借人签字']['日期'] = signature_date
signature_name, signature_date = self.get_last_page_signature(page_num='7',
......
src/common/electronic_hil_contract/get_char.py
......@@ -847,7 +847,7 @@ class Finder:
for i in range(5):
for span in ocr_results:
iou = caculate_iou(anchor, span[0])
if iou > 0 and span[1].strip() != '所购':
if iou > 0.01 and span[1].strip() != '所购':
x = get_table_info(span[0], ocr_results[key_gmjg][0], ocr_results)
y = get_table_info(span[0], ocr_results[key_sjrzje][0], ocr_results)
line = [span[1].replace('\u3000', ' '), x, y]
......@@ -1088,6 +1088,17 @@ class Finder:
# 取 Page 1 上的合同编号
contract_no = self.get_contract_no(page_num='0')
self.init_result['合同编号'] = contract_no
# 粗略判断是否是 ‘车贷分离版本’ 的合同
is_cdfl = False
for block in self.pdf_info['0']['blocks']:
if block['type'] != 0:
continue
for line in block['lines']:
for span in line['spans']:
bbox, text = span['bbox'], span['text']
if '共同承租人：' in text:
is_cdfl = True
if is_cdfl == False:
# 从第一页上取四个角色的姓名和证件号码
name, id_num, representative = self.get_role_info(role_key='承租人：', page_num='0')
if name["words"] == None:
......@@ -1124,6 +1135,23 @@ class Finder:
self.init_result['保证人3-姓名'] = name
self.init_result['保证人3-证件号码'] = id_num
self.init_result['保证人3-法定代表人或授权代表'] = representative
else:
name, id_num, representative = self.get_role_info_3_3(role_key='承租人一：', page_num='0')
self.init_result['承租人-姓名'] = name
self.init_result['承租人-证件号码'] = id_num
self.init_result['承租人-法定代表人或授权代表'] = representative
name, id_num, representative = self.get_role_info_3_3(role_key='共同承租人：', page_num='0')
self.init_result['共同承租人-姓名'] = name
self.init_result['共同承租人-证件号码'] = id_num
self.init_result['共同承租人-法定代表人或授权代表'] = representative
name, id_num, representative = self.get_role_info_3_3(role_key='保证人1：', page_num='0')
self.init_result['保证人1-姓名'] = name
self.init_result['保证人1-证件号码'] = id_num
self.init_result['保证人1-法定代表人或授权代表'] = representative
name, id_num, representative = self.get_role_info_3_3(role_key='保证人2：', page_num='0')
self.init_result['保证人2-姓名'] = name
self.init_result['保证人2-证件号码'] = id_num
self.init_result['保证人2-法定代表人或授权代表'] = representative
# 在所有页面中找正文中（第二部分 融资租赁主要条款及付款计划）的那个编号，因为存在换行的情况所以暂时不带位置输出
contract_no = self.get_contract_no_one()
self.init_result['合同编号（正文）'] = contract_no
......@@ -1157,6 +1185,7 @@ class Finder:
self.init_result['银行账户-开户行'] = bank
# 找签字页上的系列信息
# 承租人姓名、签章
if is_cdfl == False:
name = self.get_key_value(key='承租人姓名：')
electronic_signature = self.get_electronic_signature(top='承租人姓名：', bottom='保证人1姓名：')
if name["words"] == None:
......@@ -1197,11 +1226,24 @@ class Finder:
electronic_signature = self.get_electronic_signature(top='保证人2姓名：', bottom='日期：')
self.init_result['签字页-保证人2姓名'] = name
self.init_result['签字页-保证人2签章'] = electronic_signature
else:
name = self.get_key_value(key='承租人一姓名：')
electronic_signature = self.get_electronic_signature(top='承租人一姓名：', bottom='共同承租人名称：')
self.init_result['签字页-承租人姓名'] = name
self.init_result['签字页-承租人签章'] = electronic_signature
name = self.get_key_value(key='共同承租人名称：')
electronic_signature = self.get_electronic_signature(top='共同承租人名称：', bottom='保证人1姓名：')
self.init_result['签字页-共同承租人姓名'] = name
self.init_result['签字页-共同承租人签章'] = electronic_signature
name = self.get_key_value(key='保证人1姓名：')
electronic_signature = self.get_electronic_signature(top='保证人1姓名：', bottom='保证人2姓名：')
self.init_result['签字页-保证人1姓名'] = name
self.init_result['签字页-保证人1签章'] = electronic_signature
name = self.get_key_value(key='保证人2姓名：')
electronic_signature = self.get_electronic_signature(top='保证人2姓名：', bottom='保证人3姓名：')
self.init_result['签字页-保证人2姓名'] = name
self.init_result['签字页-保证人2签章'] = electronic_signature
return self.init_result
# results['is_shhz_contract'] = True
# results['pdf_info'] = self.init_result
# return results
def get_info_1(self):
if len(self.pdf_info) > 0:
......
Styling with Markdown is supported
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!