Echo/fcb_photo_review
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

项目架构调整，模型全部采用接口调用

Browse Source
This commit is contained in:
liuyebo
2024-09-25 14:46:37 +08:00
parent 7647df7d74
commit b8c1202957
25 changed files with 467 additions and 222 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
Dockerfile
View File

@@ -1,5 +1,5 @@
# 使用官方的paddle镜像作为基础
FROM registry.baidubce.com/paddlepaddle/paddle:2.6.1-gpu-cuda12.0-cudnn8.9-trt8.6
FROM python:3.10-alpine
# 设置工作目录
WORKDIR /app
@@ -15,9 +15,7 @@ ENV PYTHONUNBUFFERED=1 \
COPY requirements.txt /app/requirements.txt
COPY packages /app/packages
RUN ln -snf /usr/share/zoneinfo/$TZ /etc/localtime && echo '$TZ' > /etc/timezone \
&& pip install --no-cache-dir -r requirements.txt \
&& pip uninstall -y onnxruntime onnxruntime-gpu \
&& pip install onnxruntime-gpu==1.18.0 --extra-index-url https://aiinfra.pkgs.visualstudio.com/PublicPackages/_packaging/onnxruntime-cuda-12/pypi/simple/
&& pip install --no-cache-dir -r requirements.txt
# 将当前目录内容复制到容器的/app内
COPY . /app

2
README.md
View File

@@ -126,3 +126,5 @@ bash update.sh
2. 新增扭曲矫正功能
21. 版本号：1.14.0
1. 新增二维码识别替换高清图片功能
22. 版本号：2.0.0
1. 项目架构调整，模型全部采用接口调用

148
docker-compose.yml
View File

@@ -1,27 +1,69 @@
x-env:
&template
image: fcb_photo_review:1.14.6
x-base:
&base_template
restart: always
x-review:
&review_template
<<: *template
x-project:
&project_template
<<: *base_template
image: fcb_photo_review:1.14.6
volumes:
- ./log:/app/log
x-paddle:
&paddle_template
<<: *base_template
image: fcb_paddle:0.0.1
volumes:
- ./log:/app/log
- ./model:/app/model
services:
ocr_api:
<<: *paddle_template
build:
context: ./services/paddle_services
container_name: ocr_api
hostname: ocr_api
command: [ '-w 5 ./services/ocr_api:app --bind 0.0.0.0:5001' ]
deploy:
resources:
reservations:
devices:
- device_ids: [ '0', '1' ]
- device_ids: [ '0' ]
capabilities: [ 'gpu' ]
driver: 'nvidia'
x-mask:
&mask_template
<<: *template
volumes:
- ./log:/app/log
settlement_api:
<<: *paddle_template
container_name: settlement_api
hostname: settlement_api
command: [ '-w 5 ./services/settlement_api:app --bind 0.0.0.0:5002' ]
deploy:
resources:
reservations:
devices:
- device_ids: [ '0' ]
capabilities: [ 'gpu' ]
driver: 'nvidia'
discharge_api:
<<: *paddle_template
container_name: discharge_api
hostname: discharge_api
command: [ '-w 5 ./services/discharge_api:app --bind 0.0.0.0:5003' ]
deploy:
resources:
reservations:
devices:
- device_ids: [ '0' ]
capabilities: [ 'gpu' ]
driver: 'nvidia'
cost_api:
<<: *paddle_template
container_name: cost_api
hostname: cost_api
command: [ '-w 5 ./services/cost_api:app --bind 0.0.0.0:5004' ]
deploy:
resources:
reservations:
@@ -30,17 +72,37 @@ x-mask:
capabilities: [ 'gpu' ]
driver: 'nvidia'
services:
clas_api:
<<: *paddle_template
container_name: clas_api
hostname: clas_api
command: [ '-w 5 ./services/clas_api:app --bind 0.0.0.0:5005' ]
deploy:
resources:
reservations:
devices:
- device_ids: [ '0' ]
capabilities: [ 'gpu' ]
driver: 'nvidia'
det_api:
<<: *template
build:
context: .
<<: *paddle_template
container_name: det_api
hostname: det_api
volumes:
- ./log:/app/log
- ./model:/app/model
# command: [ 'det_api.py' ]
command: [ '-w 5 ./services/det_api:app --bind 0.0.0.0:5006' ]
deploy:
resources:
reservations:
devices:
- device_ids: [ '1' ]
capabilities: [ 'gpu' ]
driver: 'nvidia'
dewarp_api:
<<: *paddle_template
container_name: dewarp_api
hostname: dewarp_api
command: [ '-w 5 ./services/dewarp_api:app --bind 0.0.0.0:5007' ]
deploy:
resources:
reservations:
@@ -50,15 +112,23 @@ services:
driver: 'nvidia'
photo_review_1:
<<: *review_template
<<: *project_template
build:
context: .
container_name: photo_review_1
hostname: photo_review_1
depends_on:
- ocr_api
- settlement_api
- discharge_api
- cost_api
- clas_api
- det_api
- dewarp_api
command: [ 'photo_review.py', '--clean', 'True' ]
photo_review_2:
<<: *review_template
<<: *project_template
container_name: photo_review_2
hostname: photo_review_2
depends_on:
@@ -66,7 +136,7 @@ services:
command: [ 'photo_review.py' ]
photo_review_3:
<<: *review_template
<<: *project_template
container_name: photo_review_3
hostname: photo_review_3
depends_on:
@@ -74,7 +144,7 @@ services:
command: [ 'photo_review.py' ]
photo_review_4:
<<: *review_template
<<: *project_template
container_name: photo_review_4
hostname: photo_review_4
depends_on:
@@ -82,7 +152,7 @@ services:
command: [ 'photo_review.py' ]
photo_review_5:
<<: *review_template
<<: *project_template
container_name: photo_review_5
hostname: photo_review_5
depends_on:
@@ -90,33 +160,23 @@ services:
command: [ 'photo_review.py' ]
photo_mask_1:
<<: *mask_template
<<: *project_template
container_name: photo_mask_1
hostname: photo_mask_1
depends_on:
- photo_review_5
- ocr_api
- settlement_api
- discharge_api
- cost_api
- clas_api
- det_api
- dewarp_api
command: [ 'photo_mask.py', '--clean', 'True' ]
photo_mask_2:
<<: *mask_template
<<: *project_template
container_name: photo_mask_2
hostname: photo_mask_2
depends_on:
- photo_mask_1
command: [ 'photo_mask.py' ]
#
# photo_review_6:
# <<: *review_template
# container_name: photo_review_6
# hostname: photo_review_6
# depends_on:
# - photo_mask_2
# command: [ 'photo_review.py' ]
#
# photo_review_7:
# <<: *review_template
# container_name: photo_review_7
# hostname: photo_review_7
# depends_on:
# - photo_review_6
# command: [ 'photo_review.py' ]

60
photo_review/__init__.py
View File

@@ -1,6 +1,4 @@
import jieba
from paddlenlp import Taskflow
from paddleocr import PaddleOCR
'''
项目配置
@@ -11,52 +9,6 @@ PHHD_BATCH_SIZE = 10
SLEEP_MINUTES = 5
# 是否发送报错邮件
SEND_ERROR_EMAIL = True
# 是否开启布局分析
LAYOUT_ANALYSIS = False
"""
信息抽取关键词配置
"""
# 患者姓名
PATIENT_NAME = ['患者姓名']
# 入院日期
ADMISSION_DATE = ['入院日期']
# 出院日期
DISCHARGE_DATE = ['出院日期']
# 发生医疗费
MEDICAL_EXPENSES = ['费用总额']
# 个人现金支付
PERSONAL_CASH_PAYMENT = ['个人现金支付']
# 个人账户支付
PERSONAL_ACCOUNT_PAYMENT = ['个人账户支付']
# 个人自费金额
PERSONAL_FUNDED_AMOUNT = ['自费金额', '个人自费']
# 医保类别
MEDICAL_INSURANCE_TYPE = ['医保类型']
# 就诊医院
HOSPITAL = ['医院']
# 就诊科室
DEPARTMENT = ['科室']
# 主治医生
DOCTOR = ['主治医生']
# 住院号
ADMISSION_ID = ['住院号']
# 医保结算单号码
SETTLEMENT_ID = ['医保结算单号码']
# 年龄
AGE = ['年龄']
# 大写总额
UPPERCASE_MEDICAL_EXPENSES = ['大写总额']
SETTLEMENT_LIST_SCHEMA = \
(PATIENT_NAME + ADMISSION_DATE + DISCHARGE_DATE + MEDICAL_EXPENSES + PERSONAL_CASH_PAYMENT
+ PERSONAL_ACCOUNT_PAYMENT + PERSONAL_FUNDED_AMOUNT + MEDICAL_INSURANCE_TYPE + ADMISSION_ID + SETTLEMENT_ID
+ UPPERCASE_MEDICAL_EXPENSES)
DISCHARGE_RECORD_SCHEMA = \
HOSPITAL + DEPARTMENT + PATIENT_NAME + ADMISSION_DATE + DISCHARGE_DATE + DOCTOR + ADMISSION_ID + AGE
COST_LIST_SCHEMA = PATIENT_NAME + ADMISSION_DATE + DISCHARGE_DATE + MEDICAL_EXPENSES
'''
别名配置
@@ -90,15 +42,3 @@ jieba.suggest_freq(('骨', '伤'), True)
jieba.suggest_freq(('感染', '性'), True)
jieba.suggest_freq(('胆', '道'), True)
jieba.suggest_freq(('脾', '胃'), True)
'''
模型配置
'''
SETTLEMENT_IE = Taskflow('information_extraction', schema=SETTLEMENT_LIST_SCHEMA, model='uie-x-base',
task_path='model/settlement_list_model', layout_analysis=LAYOUT_ANALYSIS, precision='fp16')
DISCHARGE_IE = Taskflow('information_extraction', schema=DISCHARGE_RECORD_SCHEMA, model='uie-x-base',
task_path='model/discharge_record_model', layout_analysis=LAYOUT_ANALYSIS, precision='fp16')
COST_IE = Taskflow('information_extraction', schema=COST_LIST_SCHEMA, model='uie-x-base', device_id=1,
task_path='model/cost_list_model', layout_analysis=LAYOUT_ANALYSIS, precision='fp16')
OCR = PaddleOCR(use_angle_cls=False, show_log=False, gpu_id=1, det_db_box_thresh=0.3)

12
photo_review/auto_photo_review.py
View File

@@ -18,15 +18,15 @@ from sqlalchemy import update
from db import MysqlSession
from db.mysql import BdYljg, BdYlks, ZxIeResult, ZxIeCost, ZxIeDischarge, ZxIeSettlement, ZxPhhd, ZxPhrec
from log import HOSTNAME
from photo_review import PATIENT_NAME, ADMISSION_DATE, DISCHARGE_DATE, MEDICAL_EXPENSES, PERSONAL_CASH_PAYMENT, \
from photo_review import PHHD_BATCH_SIZE, SLEEP_MINUTES, HOSPITAL_ALIAS, HOSPITAL_FILTER, DEPARTMENT_ALIAS, \
DEPARTMENT_FILTER
from services import PATIENT_NAME, ADMISSION_DATE, DISCHARGE_DATE, MEDICAL_EXPENSES, PERSONAL_CASH_PAYMENT, \
PERSONAL_ACCOUNT_PAYMENT, PERSONAL_FUNDED_AMOUNT, MEDICAL_INSURANCE_TYPE, HOSPITAL, DEPARTMENT, DOCTOR, \
ADMISSION_ID, SETTLEMENT_ID, AGE, OCR, SETTLEMENT_IE, DISCHARGE_IE, COST_IE, PHHD_BATCH_SIZE, SLEEP_MINUTES, \
UPPERCASE_MEDICAL_EXPENSES, HOSPITAL_ALIAS, HOSPITAL_FILTER, DEPARTMENT_ALIAS, DEPARTMENT_FILTER
ADMISSION_ID, SETTLEMENT_ID, AGE, UPPERCASE_MEDICAL_EXPENSES
from ucloud import ufile
from util import image_util, common_util, html_util, model_util
from util.data_util import handle_date, handle_decimal, parse_department, handle_name, \
handle_insurance_type, handle_original_data, handle_hospital, handle_department, handle_id, handle_age, parse_money, \
parse_hospital
from util.data_util import handle_date, handle_decimal, parse_department, handle_name, handle_insurance_type, \
handle_original_data, handle_hospital, handle_department, handle_id, handle_age, parse_money, parse_hospital
# 合并信息抽取结果

17
requirements.txt
View File

@@ -1,16 +1,11 @@
numpy==1.26.4
onnxconverter-common==1.14.0
OpenCC==1.1.6
OpenCC==1.1.6 # 中文繁简转换
opencv-python==4.6.0.66
paddle2onnx==1.2.3
paddleclas==2.5.2
paddlenlp==2.6.1
paddleocr==2.7.3
pillow==10.4.0
pymysql==1.1.1
requests==2.32.3
sqlacodegen==2.3.0.post1
sqlalchemy==1.4.52
tenacity==8.5.0
ufile==3.2.9
zxing-cpp==2.2.0
sqlacodegen==2.3.0.post1 # 实体类生成
sqlalchemy==1.4.52 # ORM框架
tenacity==8.5.0 # 重试
ufile==3.2.9 # 云空间
zxing-cpp==2.2.0 # 二维码识别

34
services/__init__.py Normal file
View File

@@ -0,0 +1,34 @@
"""
信息抽取关键词配置
"""
# 患者姓名
PATIENT_NAME = ['患者姓名']
# 入院日期
ADMISSION_DATE = ['入院日期']
# 出院日期
DISCHARGE_DATE = ['出院日期']
# 发生医疗费
MEDICAL_EXPENSES = ['费用总额']
# 个人现金支付
PERSONAL_CASH_PAYMENT = ['个人现金支付']
# 个人账户支付
PERSONAL_ACCOUNT_PAYMENT = ['个人账户支付']
# 个人自费金额
PERSONAL_FUNDED_AMOUNT = ['自费金额', '个人自费']
# 医保类别
MEDICAL_INSURANCE_TYPE = ['医保类型']
# 就诊医院
HOSPITAL = ['医院']
# 就诊科室
DEPARTMENT = ['科室']
# 主治医生
DOCTOR = ['主治医生']
# 住院号
ADMISSION_ID = ['住院号']
# 医保结算单号码
SETTLEMENT_ID = ['医保结算单号码']
# 年龄
AGE = ['年龄']
# 大写总额
UPPERCASE_MEDICAL_EXPENSES = ['大写总额']

12
clas_api.py → services/clas_api.py
View File

@@ -1,26 +1,26 @@
from flask import Flask, request
from paddleclas import PaddleClas
from util.common_util import process_request
from utils import process_request
app = Flask(__name__)
CLAS = PaddleClas(model_name="text_image_orientation")
CLAS = PaddleClas(model_name='text_image_orientation')
@app.route('/clas/orientation', methods=['POST'])
@process_request
def orientation():
"""
判断图片旋转角度，逆时针旋转该角度后为正。可能值["0", "90", "180", "270"]
判断图片旋转角度，逆时针旋转该角度后为正。可能值['0', '90', '180', '270']
:return: 最有可能的两个角度
"""
img_path = request.form.get('img_path')
clas_result = CLAS.predict(input_data=img_path)
clas_result = next(clas_result)[0]
if clas_result["scores"][0] < 0.5:
if clas_result['scores'][0] < 0.5:
return ['0', '90']
return clas_result["label_names"]
return clas_result['label_names']
if __name__ == '__main__':
app.run('0.0.0.0', 5002)
app.run('0.0.0.0', 5005)

24
services/cost_api.py Normal file
View File

@@ -0,0 +1,24 @@
import json
from flask import Flask, request
from paddlenlp import Taskflow
from utils import process_request
from . import PATIENT_NAME, ADMISSION_DATE, DISCHARGE_DATE, MEDICAL_EXPENSES
app = Flask(__name__)
COST_LIST_SCHEMA = PATIENT_NAME + ADMISSION_DATE + DISCHARGE_DATE + MEDICAL_EXPENSES
COST = Taskflow('information_extraction', schema=COST_LIST_SCHEMA, model='uie-x-base',
task_path='../model/cost_list_model', layout_analysis=False, precision='fp16')
@app.route('/nlp/cost', methods=['POST'])
@process_request
def cost():
img_path = request.form.get('img_path')
layout = request.form.get('layout')
return COST({'doc': img_path, 'layout': json.loads(layout)})
if __name__ == '__main__':
app.run('0.0.0.0', 5004)

12
det_api.py → services/det_api.py
View File

@@ -3,9 +3,8 @@ import os.path
import cv2
from flask import Flask, request
from paddle_detection import detector
from util import image_util
from util.common_util import process_request
from paddle_services.paddle_detection import detector
from utils import process_request, parse_img_path
app = Flask(__name__)
@@ -14,10 +13,9 @@ app = Flask(__name__)
@process_request
def books():
img_path = request.form.get('img_path')
image = cv2.imread(img_path)
result = detector.get_book_areas(image)
result = detector.get_book_areas(img_path)
dirname, img_name, ext = image_util.parse_path(img_path)
dirname, img_name, ext = parse_img_path(img_path)
books_path = []
for i in range(len(result)):
save_path = os.path.join(dirname, img_name + '_book_' + str(i) + '.' + ext)
@@ -27,4 +25,4 @@ def books():
if __name__ == '__main__':
app.run('0.0.0.0', 5000)
app.run('0.0.0.0', 5006)

11
dewarp_api.py → services/dewarp_api.py
View File

@@ -3,9 +3,8 @@ import os
import cv2
from flask import Flask, request
from doc_dewarp import dewarp
from util import image_util
from util.common_util import process_request
from paddle_services.doc_dewarp import dewarper
from utils import process_request, parse_img_path
app = Flask(__name__)
@@ -15,12 +14,12 @@ app = Flask(__name__)
def dewarp():
img_path = request.form.get('img_path')
img = cv2.imread(img_path)
dewarped_img = dewarp.dewarp_image(img)
dirname, img_name, ext = image_util.parse_path(img_path)
dewarped_img = dewarper.dewarp_image(img)
dirname, img_name, ext = parse_img_path(img_path)
save_path = os.path.join(dirname, img_name + '_dewarped.' + ext)
cv2.imwrite(save_path, dewarped_img)
return save_path
if __name__ == '__main__':
app.run('0.0.0.0', 5001)
app.run('0.0.0.0', 5007)

26
services/discharge_api.py Normal file
View File

@@ -0,0 +1,26 @@
import json
from flask import Flask, request
from paddlenlp import Taskflow
from utils import process_request
from . import HOSPITAL, DEPARTMENT, PATIENT_NAME, ADMISSION_DATE, DISCHARGE_DATE, DOCTOR, ADMISSION_ID, AGE
app = Flask(__name__)
DISCHARGE_RECORD_SCHEMA = (
HOSPITAL + DEPARTMENT + PATIENT_NAME + ADMISSION_DATE + DISCHARGE_DATE + DOCTOR + ADMISSION_ID + AGE
)
DISCHARGE = Taskflow('information_extraction', schema=DISCHARGE_RECORD_SCHEMA, model='uie-x-base',
task_path='../model/discharge_record_model', layout_analysis=False, precision='fp16')
@app.route('/nlp/discharge', methods=['POST'])
@process_request
def discharge():
img_path = request.form.get('img_path')
layout = request.form.get('layout')
return DISCHARGE({'doc': img_path, 'layout': json.loads(layout)})
if __name__ == '__main__':
app.run('0.0.0.0', 5003)

18
services/ocr_api.py Normal file
View File

@@ -0,0 +1,18 @@
from flask import Flask, request
from paddleocr import PaddleOCR
from utils import process_request
app = Flask(__name__)
OCR = PaddleOCR(use_angle_cls=False, show_log=False, gpu_id=0, det_db_box_thresh=0.3)
@app.route('/ocr', methods=['POST'])
@process_request
def ocr():
img_path = request.form.get('img_path')
return OCR.ocr(img_path, cls=False)
if __name__ == '__main__':
app.run('0.0.0.0', 5001)

29
services/paddle_services/Dockerfile Normal file
View File

@@ -0,0 +1,29 @@
# 使用官方的paddle镜像作为基础
FROM registry.baidubce.com/paddlepaddle/paddle:2.6.1-gpu-cuda12.0-cudnn8.9-trt8.6
# 设置工作目录
WORKDIR /app
# 设置环境变量
ENV PYTHONUNBUFFERED=1 \
# 设置时区
TZ=Asia/Shanghai \
# 设置pip镜像地址，加快安装速度
PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple
# 安装依赖
COPY requirements.txt /app/requirements.txt
COPY packages /app/packages
RUN ln -snf /usr/share/zoneinfo/$TZ /etc/localtime && echo '$TZ' > /etc/timezone \
&& pip install --no-cache-dir -r requirements.txt \
&& pip uninstall -y onnxruntime onnxruntime-gpu \
&& pip install onnxruntime-gpu==1.18.0 --extra-index-url https://aiinfra.pkgs.visualstudio.com/PublicPackages/_packaging/onnxruntime-cuda-12/pypi/simple/
# 将当前目录内容复制到容器的/app内
COPY . /app
# 暴露端口
# EXPOSE 8081
# 运行api接口，具体接口在命令行或docker-compose.yml文件中定义
ENTRYPOINT ["gunicorn"]

0
paddle_detection/test_tipc/test_lite.sh → services/paddle_services/__init__.py
View File

7
services/paddle_services/doc_dewarp/__init__.py
View File

@@ -1,4 +1,7 @@
import os.path
from onnxruntime import InferenceSession
DOC_TR = InferenceSession("model/dewarp_model/doc_tr_pp.onnx",
providers=["CUDAExecutionProvider"], provider_options=[{"device_id": 0}])
MODEL_PATH = os.path.join(os.path.dirname(os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))),
'model', 'dewarp_model', 'doc_tr_pp.onnx')
DOC_TR = InferenceSession(MODEL_PATH, providers=['CUDAExecutionProvider'], provider_options=[{'device_id': 0}])

4
services/paddle_services/doc_dewarp/dewarp.py → services/paddle_services/doc_dewarp/dewarper.py
View File

@@ -11,10 +11,10 @@ def dewarp_image(image):
y = to_tensor(image)
img = np.transpose(img, (2, 0, 1))
bm = DOC_TR.run(None, {"image": img[None,]})[0]
bm = DOC_TR.run(None, {'image': img[None,]})[0]
bm = paddle.to_tensor(bm)
bm = paddle.nn.functional.interpolate(
bm, y.shape[2:], mode="bilinear", align_corners=False
bm, y.shape[2:], mode='bilinear', align_corners=False
)
bm_nhwc = np.transpose(bm, (0, 2, 3, 1))
out = paddle.nn.functional.grid_sample(y, (bm_nhwc / 288 - 0.5) * 2)

8
services/paddle_services/paddle_detection/__init__.py
View File

@@ -1,4 +1,8 @@
import os
from onnxruntime import InferenceSession
PADDLE_DET = InferenceSession("model/object_det_model/ppyoloe_plus_crn_l_80e_coco_w_nms.onnx",
providers=["CPUExecutionProvider"], provider_options=[{"device_id": 0}])
MODEL_DIR = os.path.join(os.path.dirname(os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))),
'model', 'object_det_model')
PADDLE_DET = InferenceSession(os.path.join(MODEL_DIR, 'ppyoloe_plus_crn_l_80e_coco_w_nms.onnx'),
providers=['CPUExecutionProvider'], provider_options=[{'device_id': 0}])

26
services/paddle_services/paddle_detection/detector.py
View File

@@ -1,13 +1,13 @@
import tempfile
import os.path
from collections import defaultdict
import cv2
import numpy as np
from paddle_detection import PADDLE_DET
from paddle_detection.deploy.third_engine.onnx.infer import PredictConfig
from paddle_detection.deploy.third_engine.onnx.preprocess import Compose
from util import image_util, common_util
from util import image_util
from . import PADDLE_DET, MODEL_DIR
from .deploy.third_engine.onnx.infer import PredictConfig
from .deploy.third_engine.onnx.preprocess import Compose
def predict_image(infer_config, predictor, img_path):
@@ -15,7 +15,7 @@ def predict_image(infer_config, predictor, img_path):
transforms = Compose(infer_config.preprocess_infos)
# predict image
inputs = transforms(img_path)
inputs["image"] = np.array(inputs["image"]).astype('float32')
inputs['image'] = np.array(inputs['image']).astype('float32')
inputs_name = [var.name for var in predictor.get_inputs()]
inputs = {k: inputs[k][None,] for k in inputs_name}
@@ -25,25 +25,23 @@ def predict_image(infer_config, predictor, img_path):
result = defaultdict(list)
for bbox in bboxes:
if bbox[0] > -1 and bbox[1] > infer_config.draw_threshold:
result[bbox[0]].append({"score": bbox[1], "box": bbox[2:]})
result[bbox[0]].append({'score': bbox[1], 'box': bbox[2:]})
return result
def detect_image(img_path):
infer_cfg = "model/object_det_model/infer_cfg.yml"
infer_cfg = os.path.join(MODEL_DIR, 'infer_cfg.yml')
# load infer config
infer_config = PredictConfig(infer_cfg)
return predict_image(infer_config, PADDLE_DET, img_path)
def get_book_areas(image):
with tempfile.NamedTemporaryFile(delete=False, suffix=".jpg") as temp_file:
cv2.imwrite(temp_file.name, image)
detect_result = detect_image(temp_file.name)
common_util.delete_temp_file(temp_file.name)
def get_book_areas(img_path):
detect_result = detect_image(img_path)
book_areas = detect_result[73]
result = []
image = cv2.imread(img_path)
for book_area in book_areas:
result.append(image_util.capture(image, book_area["box"]))
result.append(image_util.capture(image, book_area['box']))
return result

16
services/paddle_services/requestments.txt Normal file
View File

@@ -0,0 +1,16 @@
numpy==1.26.4
onnxconverter-common==1.14.0
OpenCC==1.1.6
opencv-python==4.6.0.66
paddle2onnx==1.2.3
paddleclas==2.5.2
paddlenlp==2.6.1
paddleocr==2.7.3
pillow==10.4.0
pymysql==1.1.1
requests==2.32.3
sqlacodegen==2.3.0.post1
sqlalchemy==1.4.52
tenacity==8.5.0
ufile==3.2.9
zxing-cpp==2.2.0

30
services/settlement_api.py Normal file
View File

@@ -0,0 +1,30 @@
import json
from flask import Flask, request
from paddlenlp import Taskflow
from utils import process_request
from . import PATIENT_NAME, ADMISSION_DATE, DISCHARGE_DATE, MEDICAL_EXPENSES, PERSONAL_CASH_PAYMENT, \
PERSONAL_ACCOUNT_PAYMENT, PERSONAL_FUNDED_AMOUNT, MEDICAL_INSURANCE_TYPE, ADMISSION_ID, SETTLEMENT_ID, \
UPPERCASE_MEDICAL_EXPENSES
app = Flask(__name__)
SETTLEMENT_LIST_SCHEMA = (
PATIENT_NAME + ADMISSION_DATE + DISCHARGE_DATE + MEDICAL_EXPENSES + PERSONAL_CASH_PAYMENT
+ PERSONAL_ACCOUNT_PAYMENT + PERSONAL_FUNDED_AMOUNT + MEDICAL_INSURANCE_TYPE + ADMISSION_ID + SETTLEMENT_ID
+ UPPERCASE_MEDICAL_EXPENSES
)
SETTLEMENT_IE = Taskflow('information_extraction', schema=SETTLEMENT_LIST_SCHEMA, model='uie-x-base',
task_path='../model/settlement_list_model', layout_analysis=False, precision='fp16')
@app.route('/nlp/settlement', methods=['POST'])
@process_request
def settlement():
img_path = request.form.get('img_path')
layout = request.form.get('layout')
return SETTLEMENT_IE({'doc': img_path, 'layout': json.loads(layout)})
if __name__ == '__main__':
app.run('0.0.0.0', 5002)

26
services/utils.py Normal file
View File

@@ -0,0 +1,26 @@
import logging
import os
from flask import jsonify
def process_request(func):
"""
api通用处理函数
"""
def wrapper(*args, **kwargs):
try:
result = func(*args, **kwargs)
return jsonify(result), 200
except Exception as e:
logging.getLogger('error').error(f'Error: {e}', exc_info=e)
return jsonify({'error': str(e)}), 500
return wrapper
def parse_img_path(img_path):
dirname = os.path.dirname(img_path)
img_name, ext = os.path.basename(img_path).rsplit('.', 1)
return dirname, img_name, ext

21
util/common_util.py
View File

@@ -2,10 +2,9 @@ import logging
import os
from datetime import datetime
from flask import jsonify
from opencc import OpenCC
from util import string_util
from util import string_util, model_util
# 获取yyyy-MM-dd HH:mm:ss格式的当前时间
@@ -37,7 +36,7 @@ def get_ocr_layout(ocr, img_path):
return True
layout = []
ocr_result = ocr.ocr(img_path, cls=False)
ocr_result = model_util.request_ocr(img_path)
ocr_result = ocr_result[0]
if not ocr_result:
return layout
@@ -218,19 +217,3 @@ def chinese_money_to_number(chinese_money_amount):
def traditional_to_simple_chinese(traditional_chinese):
converter = OpenCC('t2s')
return converter.convert(traditional_chinese)
def process_request(func):
"""
api通用处理函数
"""
def wrapper(*args, **kwargs):
try:
result = func(*args, **kwargs)
return jsonify(result), 200
except Exception as e:
logging.getLogger('error').error(f'Error: {e}')
return jsonify({'error': str(e)}), 500
return wrapper

6
util/image_util.py
View File

@@ -283,9 +283,3 @@ def save_to_local(img_url, save_path=None):
file.write(response.content)
return save_path
def parse_path(img_path):
dirname = os.path.dirname(img_path)
img_name, ext = os.path.basename(img_path).rsplit('.', 1)
return dirname, img_name, ext

104
util/model_util.py
View File

@@ -1,9 +1,93 @@
import json
import logging
import requests
from tenacity import retry, stop_after_attempt, wait_random
@retry(stop=stop_after_attempt(3), wait=wait_random(1, 3), reraise=True,
after=lambda x: logging.warning('OCR识别失败！'))
def request_ocr(img_path):
"""
请求图片OCR识别接口
:param img_path: 待识别图片路径
:return: 识别结果
"""
url = 'http://ocr_api:5001/ocr'
response = requests.post(url, {'img_path': img_path})
if response.status_code == 200:
return response.json()
else:
return None
@retry(stop=stop_after_attempt(3), wait=wait_random(1, 3), reraise=True,
after=lambda x: logging.warning('抽取基本医保结算单失败！'))
def request_discharge_info(img_path, layout):
"""
请求基本医保结算单信息抽取接口
:param img_path: 待抽取图片路径
:param layout: 图片ocr信息
:return: 抽取结果
"""
url = 'http://settlement_api:5002/nlp/settlement'
response = requests.post(url, {'img_path': img_path, 'layout': json.dumps(layout)})
if response.status_code == 200:
return response.json()
else:
return None
@retry(stop=stop_after_attempt(3), wait=wait_random(1, 3), reraise=True,
after=lambda x: logging.warning('抽取出院记录失败！'))
def request_discharge_info(img_path, layout):
"""
请求出院记录信息抽取接口
:param img_path: 待抽取图片路径
:param layout: 图片ocr信息
:return: 抽取结果
"""
url = 'http://discharge_api:5003/nlp/discharge'
response = requests.post(url, {'img_path': img_path, 'layout': json.dumps(layout)})
if response.status_code == 200:
return response.json()
else:
return None
@retry(stop=stop_after_attempt(3), wait=wait_random(1, 3), reraise=True,
after=lambda x: logging.warning('抽取费用清单失败！'))
def request_cost_info(img_path, layout):
"""
请求费用清单信息抽取接口
:param img_path: 待抽取图片路径
:param layout: 图片ocr信息
:return: 抽取结果
"""
url = 'http://cost_api:5004/nlp/cost'
response = requests.post(url, {'img_path': img_path, 'layout': json.dumps(layout)})
if response.status_code == 200:
return response.json()
else:
return None
@retry(stop=stop_after_attempt(3), wait=wait_random(1, 3), reraise=True,
after=lambda x: logging.warning('获取图片方向失败！'))
def request_image_orientation(img_path):
"""
请求图片方向分类接口
:param img_path: 待分类图片路径
:return: 最有可能的两个图片方向
"""
url = 'http://clas_api:5005/clas/orientation'
response = requests.post(url, {'img_path': img_path})
if response.status_code == 200:
return response.json()
else:
return ['0', '90']
@retry(stop=stop_after_attempt(3), wait=wait_random(1, 3), reraise=True,
after=lambda x: logging.warning('获取文档区域失败！'))
def request_book_areas(img_path):
@@ -12,7 +96,7 @@ def request_book_areas(img_path):
:param img_path: 待识别图片路径
:return: 文档图片路径列表
"""
url = 'http://det_api:5000/det/books'
url = 'http://det_api:5006/det/books'
response = requests.post(url, {'img_path': img_path})
if response.status_code == 200:
return response.json()
@@ -28,25 +112,9 @@ def request_dewarped_image(img_path):
:param img_path: 待矫正图片路径
:return: 矫正后的图片路径
"""
url = 'http://det_api:5001/dewarp'
url = 'http://127.0.0.1:5007/dewarp'
response = requests.post(url, {'img_path': img_path})
if response.status_code == 200:
return response.json()
else:
return img_path
@retry(stop=stop_after_attempt(3), wait=wait_random(1, 3), reraise=True,
after=lambda x: logging.warning('获取图片方向失败！'))
def request_image_orientation(img_path):
"""
请求图片方向分类接口
:param img_path: 待分类图片路径
:return: 最有可能的两个图片方向
"""
url = 'http://det_api:5002/clas/orientation'
response = requests.post(url, {'img_path': img_path})
if response.status_code == 200:
return response.json()
else:
return ['0', '90']