Skip to content

周伟奇 / bmw-ocr

Go to a project
Toggle navigation
Toggle navigation pinning
4d3ca7e0 by 周伟奇
Options

slice part 2

1 parent 149b0899
Inline Side-by-side
Showing 3 changed files with 219 additions and 44 deletions
src/apps/doc/consts.py
......@@ -1542,9 +1542,12 @@ BC_FIELD_ORDER_2 = (('BankName', '发卡行名称'),
IMG_PATH_KEY = 'uniq_img_path_key'
IMG_PATH_KEY_2 = 'uniq_img_path_key_2'
SECTION_IMG_PATH_KEY = 'uniq_section_img_path_key'
ALL_POSITION_KEY = 'uniq_all_position_key'
POSITION_KEY = 'uniq_position_key'
SECTION_KEY = 'uniq_section_key'
ANGLE_KEY = 'uniq_angle_key'
FIELD_POSITION_KEY = 'position'
FIELD_QUAD_KEY = 'quad'
INFO_SOURCE = ['POS', 'CMS']
......
src/apps/doc/management/commands/ocr_process.py
......@@ -373,7 +373,7 @@ class Command(BaseCommand, LoggerMixin):
license_summary.setdefault(classify, []).extend(license_data)
res_list.append((pno, ino, part_idx, consts.RES_SUCCESS))
def license2_process(self, ocr_res_2, license_summary, pid, classify, res_list, pno, ino, part_idx, img_path, do_dda, dda_id_bc_mapping, section_img_path=None):
def license2_process(self, ocr_res_2, license_summary, pid, classify, res_list, pno, ino, part_idx, img_path, do_dda, dda_id_bc_mapping, file_data):
if ocr_res_2.get('ErrorCode') in consts.SUCCESS_CODE_SET:
res_list.append((pno, ino, part_idx, consts.RES_SUCCESS))
if pid == consts.BC_PID:
......@@ -389,19 +389,51 @@ class Command(BaseCommand, LoggerMixin):
dda_id_bc_mapping.setdefault(consts.BC_FIELD, []).append((bc_no, img_path))
else:
# 营业执照等
for result_dict in ocr_res_2.get('ResultList', []):
pre, suf = os.path.splitext(img_path)
src_section_img_path = img_path if file_data is None else '{0}_{1}{2}'.format(pre, part_idx, suf)
is_save = False
for res_idx, result_dict in enumerate(ocr_res_2.get('ResultList', [])):
image_data = result_dict.get('image_data', '')
if len(image_data) > 0:
position = {}
angle = 0
section_img_path = '{0}_{1}_{2}{3}'.format(pre, part_idx, res_idx, suf)
try:
with open(section_img_path, "wb") as fh:
fh.write(base64.b64decode(image_data.encode()))
except Exception as e:
self.online_log.warn(
'{0} [section img save failed] [img_path={1}]'
' [part_idx={2}] [res_idx={3}]'.format(self.log_base, img_path, part_idx, res_idx))
else:
is_save = True
section_img_path = src_section_img_path
position = result_dict.get('position', {})
angle = result_dict.get('angle', 0)
res_dict = {}
position_dict = {}
for field_dict in result_dict.get('FieldList', []):
res_dict[field_dict.get('chn_key', '')] = field_dict.get('value', '')
position_dict[field_dict.get('chn_key', '')] = field_dict.get('position', [])
position_dict[consts.SECTION_KEY] = position
position_dict[field_dict.get('chn_key', '')] = {
consts.FIELD_POSITION_KEY: field_dict.get('position', {}),
consts.FIELD_QUAD_KEY: field_dict.get('quad', []),
}
position_dict[consts.POSITION_KEY] = position
position_dict[consts.ANGLE_KEY] = angle
res_dict[consts.IMG_PATH_KEY] = section_img_path if isinstance(section_img_path, str) else img_path
res_dict[consts.POSITION_KEY] = position_dict
res_dict[consts.IMG_PATH_KEY] = img_path
res_dict[consts.SECTION_IMG_PATH_KEY] = section_img_path
res_dict[consts.ALL_POSITION_KEY] = position_dict
license_summary.setdefault(classify, []).append(res_dict)
if is_save and file_data is not None:
try:
with open(src_section_img_path, "wb") as fh:
fh.write(base64.b64decode(file_data.encode()))
except Exception as e:
self.online_log.warn(
'{0} [section img save failed] [img_path={1}]'
' [part_idx={2}]'.format(self.log_base, img_path, part_idx))
else:
res_list.append((pno, ino, part_idx, consts.RES_FAILED_2))
......@@ -980,22 +1012,9 @@ class Command(BaseCommand, LoggerMixin):
name = '无'
ocr_2_res['Name'] = name
section_img_path = None
try:
if ocr_2_res.get('ErrorCode') in consts.SUCCESS_CODE_SET:
pre, suf = os.path.splitext(img_path)
section_img_path = '{0}_{1}{2}'.format(pre, part_idx, suf)
with open(section_img_path, "wb") as fh:
fh.write(base64.b64decode(file_data.encode()))
except Exception as e:
self.online_log.warn(
'{0} [section img save failed] [img_path={1}]'
' [part_idx={2}]'.format(self.log_base, img_path, part_idx))
self.license2_process(ocr_2_res, license_summary, pid, classify,
res_list, pno, ino, part_idx, img_path,
do_dda, dda_id_bc_mapping,
section_img_path=section_img_path)
do_dda, dda_id_bc_mapping, file_data=ocr_data.get('section_img'))
break
else:
res_list.append((pno, ino, part_idx, consts.RES_FAILED_2))
......
src/celery_compare/tasks.py
import json
import os
import cv2
import time
import logging
import traceback
import numpy as np
from datetime import datetime, timedelta
from collections import OrderedDict
from . import app
......@@ -40,6 +43,79 @@ empty_error_type = 1000
des_key = conf.CMS_DES_KEY
def rotate_bound(image, angle):
# grab the dimensions of the image and then determine the
# center
(h, w) = image.shape[:2]
(cX, cY) = (w // 2, h // 2)
# grab the rotation matrix (applying the negative of the
# angle to rotate clockwise), then grab the sine and cosine
# (i.e., the rotation components of the matrix)
M = cv2.getRotationMatrix2D((cX, cY), angle, 1.0)
cos = np.abs(M[0, 0])
sin = np.abs(M[0, 1])
# compute the new bounding dimensions of the image
nW = int((h * sin) + (w * cos))
nH = int((h * cos) + (w * sin))
# adjust the rotation matrix to take into account translation
M[0, 2] += (nW / 2) - cX
M[1, 2] += (nH / 2) - cY
# perform the actual rotation and return the image
return cv2.warpAffine(image, M, (nW, nH))
def build_coordinates(section_position_dict):
if isinstance(section_position_dict, dict):
h_min = section_position_dict.get('top', 0)
w_min = section_position_dict.get('left', 0)
h_max = h_min + section_position_dict.get('height', 0)
w_max = w_min + section_position_dict.get('width', 0)
if h_max > h_min and w_max > w_min:
return True, (h_min, h_max, w_min, w_max)
else:
return False, ()
return False, ()
def field_build_coordinates(field_position_info):
field_position_dict = field_position_info.get(consts.FIELD_POSITION_KEY, {})
field_quad_list = field_position_info.get(consts.FIELD_QUAD_KEY, [])
if isinstance(field_quad_list, list) and len(field_quad_list) == 8:
w_list = [field_quad_list[0], field_quad_list[2], field_quad_list[4], field_quad_list[6]]
h_list = [field_quad_list[1], field_quad_list[3], field_quad_list[5], field_quad_list[7]]
h_min = min(h_list)
h_max = max(h_list)
w_min = min(w_list)
w_max = max(w_list)
if h_max > h_min and w_max > w_min:
return True, (h_min, h_max, w_min, w_max)
if isinstance(field_position_dict, dict):
h_min = field_position_dict.get('top', 0)
w_min = field_position_dict.get('left', 0)
h_max = h_min + field_position_dict.get('height', 0)
w_max = w_min + field_position_dict.get('width', 0)
if h_max > h_min and w_max > w_min:
return True, (h_min, h_max, w_min, w_max)
else:
return False, ()
return False, ()
def img_process(section_img_path, section_position, section_angle):
image = cv2.imread(section_img_path)
is_valid, coord_tuple = build_coordinates(section_position)
if is_valid:
image = image[coord_tuple[0]:coord_tuple[1], coord_tuple[2]:coord_tuple[3], :]
if isinstance(section_angle, int) or isinstance(section_angle, float):
if section_angle != 0:
return rotate_bound(image, section_angle)
return image
class FakePOS:
def __init__(self,
......@@ -289,6 +365,8 @@ def ca_compare_license(license_en, ocr_res_dict, field_list):
is_find = False
result_field_list = []
section_img_info = dict()
field_img_path_dict = dict()
ocr_res_str = ocr_res_dict.get(ocr_field)
if ocr_res_str is not None:
ocr_res_list = json.loads(ocr_res_str)
......@@ -323,6 +401,8 @@ def ca_compare_license(license_en, ocr_res_dict, field_list):
break
is_find = True
section_img_info[consts.SECTION_IMG_PATH_KEY] = ocr_res_list[res_idx].get(consts.SECTION_IMG_PATH_KEY, '')
section_img_info[consts.ALL_POSITION_KEY] = ocr_res_list[res_idx].get(consts.ALL_POSITION_KEY, {})
# 过期期限特殊处理
if special_expiry_date and name == 'idExpiryDate' and result == consts.RESULT_N:
......@@ -341,13 +421,48 @@ def ca_compare_license(license_en, ocr_res_dict, field_list):
else:
img_path = ocr_res_list[res_idx].get(consts.IMG_PATH_KEY, '') if result == consts.RESULT_N else empty_str
error_type = empty_error_type if result == consts.RESULT_Y else ErrorType.OCR.value
result_field_list.append((name, value, result, ocr_str, position_img_path, img_path, error_type))
result_field_list.append((name, value, result, ocr_str, img_path, error_type))
if not is_find:
for name, value in field_list:
result_field_list.append((name, value, consts.RESULT_N, empty_str, empty_str, empty_str, ErrorType.NF.value))
result_field_list.append((name, value, consts.RESULT_N, empty_str, empty_str, ErrorType.NF.value))
if is_find:
section_img_path = section_img_info.get(consts.SECTION_IMG_PATH_KEY, '')
if os.path.exists(section_img_path):
failed_field = []
base_img_path = empty_str
for name, _, result, _, img_path, _ in result_field_list:
if result == consts.RESULT_N:
failed_field.append(name)
if base_img_path == empty_str:
base_img_path = img_path
if len(failed_field) > 0:
info = section_img_info.get(consts.ALL_POSITION_KEY, {})
section_position = info.get(consts.POSITION_KEY, {})
section_angle = info.get(consts.ANGLE_KEY, 0)
try:
last_img = img_process(section_img_path, section_position, section_angle)
except Exception as e:
for field in failed_field:
field_img_path_dict[field] = base_img_path
else:
pre, suf = os.path.splitext(section_img_path)
for field in failed_field:
try:
res_field = compare_logic[field][0]
is_valid, coord_tuple = field_build_coordinates(info.get(res_field, {}))
if is_valid:
save_path = '{0}_{1}{2}'.format(pre, field, suf)
field_img = last_img[coord_tuple[0]:coord_tuple[1], coord_tuple[2]:coord_tuple[3], :]
cv2.imwrite(save_path, field_img)
field_img_path_dict[field] = save_path
else:
field_img_path_dict[field] = base_img_path
except Exception as e:
field_img_path_dict[field] = base_img_path
return result_field_list
return result_field_list, field_img_path_dict
def ca_compare_process(compare_info, ocr_res_dict):
......@@ -363,8 +478,8 @@ def ca_compare_process(compare_info, ocr_res_dict):
for idx, license_list in info_value.items():
for license_dict in license_list:
for license_en, field_list in license_dict.items():
result_field_list = ca_compare_license(license_en, ocr_res_dict, field_list)
for name, value, result, ocr_str, position_img_path, img_path, error_type in result_field_list:
result_field_list, field_img_path_dict = ca_compare_license(license_en, ocr_res_dict, field_list)
for name, value, result, ocr_str, img_path, error_type in result_field_list:
total_fields += 1
if result == consts.RESULT_N:
failed_count += 1
......@@ -377,15 +492,15 @@ def ca_compare_process(compare_info, ocr_res_dict):
consts.HEAD_LIST[4]: value,
consts.HEAD_LIST[5]: ocr_str,
consts.HEAD_LIST[6]: result,
consts.HEAD_LIST[7]: position_img_path,
consts.HEAD_LIST[7]: field_img_path_dict.get(name, empty_str),
consts.HEAD_LIST[8]: img_path,
consts.HEAD_LIST[9]: error_type,
}
)
else:
for license_en, field_list in info_value.items():
result_field_list = ca_compare_license(license_en, ocr_res_dict, field_list)
for name, value, result, ocr_str, position_img_path, img_path, error_type in result_field_list:
result_field_list, field_img_path_dict = ca_compare_license(license_en, ocr_res_dict, field_list)
for name, value, result, ocr_str, img_path, error_type in result_field_list:
total_fields += 1
if result == consts.RESULT_N:
failed_count += 1
......@@ -398,7 +513,7 @@ def ca_compare_process(compare_info, ocr_res_dict):
consts.HEAD_LIST[4]: value,
consts.HEAD_LIST[5]: ocr_str,
consts.HEAD_LIST[6]: result,
consts.HEAD_LIST[7]: position_img_path,
consts.HEAD_LIST[7]: field_img_path_dict.get(name, empty_str),
consts.HEAD_LIST[8]: img_path,
consts.HEAD_LIST[9]: error_type,
}
......@@ -1073,6 +1188,8 @@ def se_compare_license(license_en, ocr_res_dict, field_list):
is_find = False
no_ocr_result = False
result_field_list = []
section_img_info = dict()
field_img_path_dict = dict()
ocr_res_str = ocr_res_dict.get(ocr_field)
if ocr_res_str is not None:
ocr_res_list = json.loads(ocr_res_str)
......@@ -1104,6 +1221,8 @@ def se_compare_license(license_en, ocr_res_dict, field_list):
break
is_find = True
section_img_info[consts.SECTION_IMG_PATH_KEY] = ocr_res_list[res_idx].get(consts.SECTION_IMG_PATH_KEY, '')
section_img_info[consts.ALL_POSITION_KEY] = ocr_res_list[res_idx].get(consts.ALL_POSITION_KEY, {})
# 过期期限特殊处理
if special_expiry_date and name == 'idExpiryDate' and result == consts.RESULT_N:
......@@ -1116,15 +1235,50 @@ def se_compare_license(license_en, ocr_res_dict, field_list):
img_path = ocr_res_list[res_idx].get(consts.IMG_PATH_KEY, '') if result == consts.RESULT_N else empty_str
error_type = empty_error_type if result == consts.RESULT_Y else ErrorType.OCR.value
result_field_list.append((name, value, result, ocr_str, position_img_path, img_path, error_type))
result_field_list.append((name, value, result, ocr_str, img_path, error_type))
else:
no_ocr_result = True
if not is_find:
for name, value in field_list:
result_field_list.append((name, value, consts.RESULT_N, empty_str, empty_str, empty_str, ErrorType.NF.value))
result_field_list.append((name, value, consts.RESULT_N, empty_str, empty_str, ErrorType.NF.value))
if is_find:
section_img_path = section_img_info.get(consts.SECTION_IMG_PATH_KEY, '')
if os.path.exists(section_img_path):
failed_field = []
base_img_path = empty_str
for name, _, result, _, img_path, _ in result_field_list:
if result == consts.RESULT_N:
failed_field.append(name)
if base_img_path == empty_str:
base_img_path = img_path
if len(failed_field) > 0:
info = section_img_info.get(consts.ALL_POSITION_KEY, {})
section_position = info.get(consts.POSITION_KEY, {})
section_angle = info.get(consts.ANGLE_KEY, 0)
try:
last_img = img_process(section_img_path, section_position, section_angle)
except Exception as e:
for field in failed_field:
field_img_path_dict[field] = base_img_path
else:
pre, suf = os.path.splitext(section_img_path)
for field in failed_field:
try:
res_field = compare_logic[field][0]
is_valid, coord_tuple = field_build_coordinates(info.get(res_field, {}))
if is_valid:
save_path = '{0}_{1}{2}'.format(pre, field, suf)
field_img = last_img[coord_tuple[0]:coord_tuple[1], coord_tuple[2]:coord_tuple[3], :]
cv2.imwrite(save_path, field_img)
field_img_path_dict[field] = save_path
else:
field_img_path_dict[field] = base_img_path
except Exception as e:
field_img_path_dict[field] = base_img_path
return result_field_list, no_ocr_result
return result_field_list, no_ocr_result, field_img_path_dict
def se_mvc34_compare(license_en, ocr_res_dict, field_list):
......@@ -1184,14 +1338,13 @@ def se_mvc34_compare(license_en, ocr_res_dict, field_list):
result = getattr(cp, compare_logic[name][1])(value, ocr_str, **compare_logic[name][2])
img_path = ocr_res.get(consts.IMG_PATH_KEY, '') if result == consts.RESULT_N else empty_str
error_type = empty_error_type if result == consts.RESULT_Y else ErrorType.OCR.value
position_img_path = empty_str
result_field_list.append((name, value, result, ocr_str, position_img_path, img_path, error_type))
result_field_list.append((name, value, result, ocr_str, img_path, error_type))
if not is_find:
for name, value in field_list:
result_field_list.append((name, value, consts.RESULT_N, empty_str, empty_str, empty_str, ErrorType.NF.value))
result_field_list.append((name, value, consts.RESULT_N, empty_str, empty_str, ErrorType.NF.value))
return result_field_list
return result_field_list, dict()
def se_compare_process(compare_info, ocr_res_dict):
......@@ -1211,8 +1364,8 @@ def se_compare_process(compare_info, ocr_res_dict):
for license_dict in license_list:
for license_en, field_list in license_dict.items():
failure_field = []
result_field_list, no_ocr_result = se_compare_license(license_en, ocr_res_dict, field_list)
for name, value, result, ocr_str, position_img_path, img_path, error_type in result_field_list:
result_field_list, no_ocr_result, field_img_path_dict = se_compare_license(license_en, ocr_res_dict, field_list)
for name, value, result, ocr_str, img_path, error_type in result_field_list:
if license_en not in consts.SKIP_CARD or not no_ocr_result:
total_fields += 1
if result == consts.RESULT_N:
......@@ -1228,7 +1381,7 @@ def se_compare_process(compare_info, ocr_res_dict):
consts.HEAD_LIST[4]: value,
consts.HEAD_LIST[5]: ocr_str,
consts.HEAD_LIST[6]: result,
consts.HEAD_LIST[7]: position_img_path,
consts.HEAD_LIST[7]: field_img_path_dict.get(name, empty_str),
consts.HEAD_LIST[8]: img_path,
consts.HEAD_LIST[9]: error_type,
}
......@@ -1239,10 +1392,10 @@ def se_compare_process(compare_info, ocr_res_dict):
for license_en, field_list in info_value.items():
failure_field = []
if license_en == consts.MVC34_EN:
result_field_list = se_mvc34_compare(license_en, ocr_res_dict, field_list)
result_field_list, field_img_path_dict = se_mvc34_compare(license_en, ocr_res_dict, field_list)
else:
result_field_list, _ = se_compare_license(license_en, ocr_res_dict, field_list)
for name, value, result, ocr_str, position_img_path, img_path, error_type in result_field_list:
result_field_list, _, field_img_path_dict = se_compare_license(license_en, ocr_res_dict, field_list)
for name, value, result, ocr_str, img_path, error_type in result_field_list:
total_fields += 1
if result == consts.RESULT_N:
failed_count += 1
......@@ -1257,7 +1410,7 @@ def se_compare_process(compare_info, ocr_res_dict):
consts.HEAD_LIST[4]: value,
consts.HEAD_LIST[5]: ocr_str,
consts.HEAD_LIST[6]: result,
consts.HEAD_LIST[7]: position_img_path,
consts.HEAD_LIST[7]: field_img_path_dict.get(name, empty_str),
consts.HEAD_LIST[8]: img_path,
consts.HEAD_LIST[9]: error_type,
}
......
Styling with Markdown is supported
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!