Skip to content

周伟奇 / bmw-ocr

Go to a project
Toggle navigation
Toggle navigation pinning
086507c6 by 周伟奇
Options

rebuild result

1 parent 5c08056d
Inline Side-by-side
Showing 2 changed files with 146 additions and 31 deletions
src/apps/doc/consts.py
......@@ -658,6 +658,13 @@ CLASSIFY_HEADER_LIST = [
OTHER_CLASSIFY = 2
# 身份证
IC_MAP_0 = (('姓名', '姓名'),
('公民身份号码', '公民身份号码'),
('出生年月', '出生'),
('住址', '住址'),
('性别', '性别'),
('民族', '民族'),)
IC_MAP_1 = (('签发机关', '签发机关'),)
IC_CN_NAME = '身份证'
IC_CLASSIFY = 33
IC_TURE_OR_FALSE = '真伪'
......@@ -670,6 +677,12 @@ IC_FIELD_ORDER_0 = ((IC_TURE_OR_FALSE, '身份证'),
('民族', '民族'),)
IC_FIELD_ORDER_1 = ((IC_TURE_OR_FALSE, '身份证'), ('有效期限', '有效期限'), ('签发机关', '签发机关'),)
# 居住证
RP_MAP_0 = (('姓名', '姓名'),
('公民身份号码', '公民身份号码'),
('出生年月', '出生'),
('住址', '住址'),
('性别', '性别'),)
RP_MAP_1 = (('签发机关', '签发机关'), ('通行证号码', '通行证号码'))
RP_CN_NAME = '居住证'
RP_CLASSIFY = 10087
RP_FIELD_ORDER_0 = (('姓名', '姓名'),
......
src/apps/doc/management/commands/ocr_process.py
......@@ -246,6 +246,59 @@ class Command(BaseCommand, LoggerMixin):
# contract_result.setdefault(page_num, []).append(rebuild_page_info)
contract_result.setdefault(classify, dict()).setdefault(page_num, []).append(rebuild_page_info)
# def rebuild_result(self, ocr_data, classify, img_path):
# license_data = ocr_data.get('data')
# if not license_data:
# return
# if classify == consts.IC_CLASSIFY:
# rebuild_data_dict = {}
# card_type = license_data.get('type', '')
# is_ic = card_type.startswith('身份证')
# is_info_side = card_type.endswith('信息面')
# rebuild_data_dict['类别'] = '0' if is_ic else '1'
# if is_ic:
# field_map = consts.IC_MAP_0 if is_info_side else consts.IC_MAP_1
# else:
# field_map = consts.RP_MAP_0 if is_info_side else consts.RP_MAP_1
# for write_field, search_field in field_map:
# rebuild_data_dict[write_field] = license_data.get('words_result', {}).get(search_field, {}).get('words', '')
# if not is_info_side:
# start_time = license_data.get('words_result', {}).get('签发日期', {}).get('words', '')
# end_time = license_data.get('words_result', {}).get('失效日期', {}).get('words', '')
# rebuild_data_dict['有效期限'] = '{0}-{1}'.format(start_time, end_time)
# return [rebuild_data_dict]
# elif classify == consts.MVC_CLASSIFY:
# # license_data[consts.IMG_PATH_KEY] = img_path
# rebuild_data_dict = {}
# mvc_page = license_data.pop('page', 'VehicleRCI')
# mvc_res = license_data.pop('results', {})
# if mvc_page == 'VehicleRegArea':
# rebuild_data_dict['机动车登记证书编号'] = mvc_res.get('机动车登记证书编号', {}).get('words', '')
# for register_info in mvc_res.get('登记信息', []):
# register_info.pop('register_type', None)
# register_info.pop('register_type_name', None)
# for cn_key, detail_dict in register_info.items():
# rebuild_data_dict.setdefault(cn_key, []).append(
# detail_dict.get('words', ''))
# else:
# for cn_key, detail_dict in mvc_res.items():
# rebuild_data_dict[cn_key] = detail_dict.get('words', '')
# del mvc_res
# return [rebuild_data_dict]
# elif classify == consts.MVI_CLASSIFY:
# rebuild_data_dict = {}
# mvi_res = license_data.pop('result', {})
# for en_key, detail_dict in mvi_res.items():
# rebuild_data_dict[detail_dict.get('chinese_key', '')] = detail_dict.get('words', '')
# return [rebuild_data_dict]
# elif classify == consts.UCI_CLASSIFY:
# rebuild_data_dict = {}
# mvi_res = license_data.pop('result', {})
# for en_key, detail_dict in mvi_res.items():
# rebuild_data_dict[detail_dict.get('chinese_key', '')] = detail_dict.get('words', '')
# return [rebuild_data_dict]
def license1_process(self, ocr_data, license_summary, classify, res_list, pno, ino, part_idx, img_path, do_dda, dda_id_bc_mapping):
# 类别：'0'身份证， '1'居住证
......@@ -253,7 +306,8 @@ class Command(BaseCommand, LoggerMixin):
if not license_data:
res_list.append((pno, ino, part_idx, consts.RES_SUCCESS_EMPTY))
return
if classify == consts.INSURANCE_CLASSIFY: # 保单处理
# 保单
if classify == consts.INSURANCE_CLASSIFY:
product_result = ['', '', '']
for product in license_data.get('result', {}).get('productList', []):
name = product.get('name', {}).get('words', '')
......@@ -281,7 +335,8 @@ class Command(BaseCommand, LoggerMixin):
'特别约定第一受益人': special,
}
license_summary.setdefault(classify, []).append(insurance_ocr_result)
elif classify == consts.DDA_CLASSIFY: # DDA处理
# DDA
elif classify == consts.DDA_CLASSIFY:
pro = ocr_data.get('confidence', 0)
if pro < consts.DDA_PRO_MIN:
res_list.append((pno, ino, part_idx, consts.RES_SUCCESS_EMPTY))
......@@ -291,40 +346,80 @@ class Command(BaseCommand, LoggerMixin):
dda_ocr_result[consts.DDA_PRO] = pro
dda_ocr_result[consts.IMG_PATH_KEY] = img_path
license_summary.setdefault(classify, []).append(dda_ocr_result)
elif classify == consts.HMH_CLASSIFY: # 抵押登记豁免函
hmh_ocr_result = {key: value.get('words', '') for key, value in license_data.items()}
# 抵押登记豁免函
elif classify == consts.HMH_CLASSIFY:
hmh_ocr_result = {key: value.get('words', '') for key, value in license_data.get('words_result', {}).items()}
license_summary.setdefault(classify, []).append(hmh_ocr_result)
elif classify == consts.JYPZ_CLASSIFY: # 二手车交易凭证
# 二手车交易凭证
elif classify == consts.JYPZ_CLASSIFY:
jypz_ocr_result = {key: value.get('words', '') for key, value in license_data.get('result', {}).items()}
license_summary.setdefault(classify, []).append(jypz_ocr_result)
elif classify == consts.MVC_CLASSIFY: # 车辆登记证 3/4页结果整合
for mvc_dict in license_data:
mvc_dict[consts.IMG_PATH_KEY] = img_path
try:
mvc_page = mvc_dict.pop('page')
except Exception as e:
pass
else:
if mvc_page == 'VehicleRegArea':
mvc_res = mvc_dict.pop('results', {})
mvc_dict['机动车登记证书编号'] = mvc_res.get('register_no', {}).get('words', '')
for register_info in mvc_res.get('register_info', []):
for detail_dict in register_info.get('details', {}).values():
mvc_dict.setdefault(detail_dict.get('chinese_key', '未知'), []).append(
detail_dict.get('words', ''))
del mvc_res
license_summary.setdefault(classify, []).extend(license_data)
# 车辆登记证 3/4页结果整合
elif classify == consts.MVC_CLASSIFY:
rebuild_data_dict = {}
rebuild_data_dict[consts.IMG_PATH_KEY] = img_path
mvc_page = license_data.pop('page', 'VehicleRCI')
mvc_res = license_data.pop('results', {})
if mvc_page == 'VehicleRegArea':
rebuild_data_dict['机动车登记证书编号'] = mvc_res.get('机动车登记证书编号', {}).get('words', '')
for register_info in mvc_res.get('登记信息', []):
register_info.pop('register_type', None)
register_info.pop('register_type_name', None)
for cn_key, detail_dict in register_info.items():
rebuild_data_dict.setdefault(cn_key, []).append(
detail_dict.get('words', ''))
else:
for cn_key, detail_dict in mvc_res.items():
rebuild_data_dict[cn_key] = detail_dict.get('words', '')
del mvc_res
license_summary.setdefault(classify, []).append(rebuild_data_dict)
# for mvc_dict in license_data:
# mvc_dict[consts.IMG_PATH_KEY] = img_path
# try:
# mvc_page = mvc_dict.pop('page')
# except Exception as e:
# pass
# else:
# if mvc_page == 'VehicleRegArea':
# mvc_res = mvc_dict.pop('results', {})
# mvc_dict['机动车登记证书编号'] = mvc_res.get('register_no', {}).get('words', '')
# for register_info in mvc_res.get('register_info', []):
# for detail_dict in register_info.get('details', {}).values():
# mvc_dict.setdefault(detail_dict.get('chinese_key', '未知'), []).append(
# detail_dict.get('words', ''))
# del mvc_res
# license_summary.setdefault(classify, []).extend(license_data)
# 身份证真伪
elif classify == consts.IC_CLASSIFY:
id_card_dict = {}
card_type = license_data.get('type', '')
is_ic = card_type.startswith('身份证')
is_info_side = card_type.endswith('信息面')
id_card_dict['类别'] = '0' if is_ic else '1'
if is_ic:
field_map = consts.IC_MAP_0 if is_info_side else consts.IC_MAP_1
else:
field_map = consts.RP_MAP_0 if is_info_side else consts.RP_MAP_1
for write_field, search_field in field_map:
id_card_dict[write_field] = license_data.get('words_result', {}).get(search_field, {}).get('words', '')
if not is_info_side:
start_time = license_data.get('words_result', {}).get('签发日期', {}).get('words', '')
end_time = license_data.get('words_result', {}).get('失效日期', {}).get('words', '')
id_card_dict['有效期限'] = '{0}-{1}'.format(start_time, end_time)
elif classify == consts.IC_CLASSIFY: # 身份证真伪
for id_card_dict in license_data:
if '有效期限' in id_card_dict:
id_card_dict[consts.IMG_PATH_KEY_2] = img_path
else:
id_card_dict[consts.IMG_PATH_KEY] = img_path
if not is_info_side:
id_card_dict[consts.IMG_PATH_KEY_2] = img_path
else:
id_card_dict[consts.IMG_PATH_KEY] = img_path
if is_ic:
try:
base64_img = id_card_dict.pop('base64_img')
base64_img = license_data.pop('base64_img')
except Exception as e:
continue
pass
else:
card_type = -1
json_data_4 = {
......@@ -371,7 +466,14 @@ class Command(BaseCommand, LoggerMixin):
ic_id = id_card_dict[consts.IC_KEY_FIELD[1]].strip()
if len(ic_name) > 0 and len(ic_id) > 0:
dda_id_bc_mapping.setdefault(consts.IC_FIELD, []).append((ic_name, ic_id, img_path))
license_summary.setdefault(classify, []).extend(license_data)
license_summary.setdefault(classify, []).append(id_card_dict)
elif classify == consts.MVI_CLASSIFY or classify == consts.UCI_CLASSIFY:
rebuild_data_dict = {}
mvi_res = license_data.pop('result', {})
for en_key, detail_dict in mvi_res.items():
rebuild_data_dict[detail_dict.get('chinese_key', '')] = detail_dict.get('words', '')
license_summary.setdefault(classify, []).append(rebuild_data_dict)
# 其他
else:
for res_dict in license_data:
res_dict[consts.IMG_PATH_KEY] = img_path
......
Styling with Markdown is supported
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!