四、OCR-docker部署最新版PaddleServing和PaddleOCR，服务端，客户端，dockerfile部署-pipeline python部署，GPU运行

1、服务端部署

1.1 dockerfile编写

FROM registry.baidubce.com/paddlepaddle/serving:0.9.0-cuda10.1-cudnn7-devel
#FROM quadtalent-docker.pkg.coding.net/smart-city/city-wise/call-12345-server-prd:1.0.0.0330


COPY . /deploy
WORKDIR /deploy

# Install requirements
RUN pip config set global.index-url https://mirror.baidu.com/pypi/simple \
    && python3.7 -m pip install --upgrade setuptools \
    && python3.7 -m pip install --upgrade pip \
    && python3.7 -m pip install paddlepaddle-gpu==2.3.0.post101 \
    -f https://www.paddlepaddle.org.cn/whl/linux/mkl/avx/stable.html \
    && pip3.7 install -r requirements.txt



ENTRYPOINT python3.7 web_service.py

1.2、制作镜像，启动容器指令

nvidia-docker build -t xx/ocr-python-server-qa:1.0.0.0630 .

nvidia-docker run -p 18091:18091 --name ocr-python-server-qa -d xx/ocr-python-server-qa:1.0.0.0630

nvidia-docker exec -it ocr-python-server-qa /bin/bash

docker logs -f ocr-python-server-qa

1.3、配置文件

#rpc端口, rpc_port和http_port不允许同时为空。当rpc_port为空且http_port不为空时，会自动将rpc_port设置为http_port+1
rpc_port: 18091

#http端口, rpc_port和http_port不允许同时为空。当rpc_port可用且http_port为空时，不自动生成http_port
http_port: 9998

#worker_num, 最大并发数。当build_dag_each_worker=True时, 框架会创建worker_num个进程，每个进程内构建grpcSever和DAG
##当build_dag_each_worker=False时，框架会设置主线程grpc线程池的max_workers=worker_num
worker_num: 10

#build_dag_each_worker, False，框架在进程内创建一条DAG；True，框架会每个进程内创建多个独立的DAG
build_dag_each_worker: False

dag:
    #op资源类型, True, 为线程模型；False，为进程模型
    is_thread_op: False

    #重试次数
    retry: 10

    #使用性能分析, True，生成Timeline性能数据，对性能有一定影响；False为不使用
    use_profile: True
    
    tracer:
        interval_s: 10
op:
    det:
        #并发数，is_thread_op=True时，为线程并发；否则为进程并发
        concurrency: 2

        #当op配置没有server_endpoints时，从local_service_conf读取本地服务配置
        local_service_conf:
            #client类型，包括brpc, grpc和local_predictor.local_predictor不启动Serving服务，进程内预测
            client_type: local_predictor

            #det模型路径
            model_config: ./ppocr_det_v3_serving

            #Fetch结果列表，以client_config中fetch_var的alias_name为准，不设置默认取全部输出变量
            #fetch_list: ["sigmoid_0.tmp_0"]

            #计算硬件ID，当devices为""或不写时为CPU预测；当devices为"0", "0,1,2"时为GPU预测，表示使用的GPU卡
            devices: "0"
            
            ir_optim: True
    rec:
        #并发数，is_thread_op=True时，为线程并发；否则为进程并发
        concurrency: 1

        #超时时间, 单位ms
        timeout: -1
 
        #Serving交互重试次数，默认不重试
        retry: 1

        #当op配置没有server_endpoints时，从local_service_conf读取本地服务配置
        local_service_conf:

            #client类型，包括brpc, grpc和local_predictor。local_predictor不启动Serving服务，进程内预测
            client_type: local_predictor

            #rec模型路径
            model_config: ./ppocr_rec_v3_serving

            #Fetch结果列表，以client_config中fetch_var的alias_name为准, 不设置默认取全部输出变量
            #fetch_list: 

            #计算硬件ID，当devices为""或不写时为CPU预测；当devices为"0", "0,1,2"时为GPU预测，表示使用的GPU卡
            devices: "0"
            
            ir_optim: True

1.4、启动文件

# Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from paddle_serving_server.web_service import WebService, Op

import logging
import numpy as np
import copy
import cv2
import base64
# from paddle_serving_app.reader import OCRReader
from ocr_reader import OCRReader, DetResizeForTest
from paddle_serving_app.reader import Sequential, ResizeByFactor
from paddle_serving_app.reader import Div, Normalize, Transpose
from paddle_serving_app.reader import DBPostProcess, FilterBoxes, GetRotateCropImage, SortedBoxes

_LOGGER = logging.getLogger()


class DetOp(Op):
    def init_op(self):
        self.det_preprocess = Sequential([
            DetResizeForTest(), Div(255),
            Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225]), Transpose(
                (2, 0, 1))
        ])
        self.filter_func = FilterBoxes(10, 10)
        self.post_func = DBPostProcess({
    
    
            "thresh": 0.3,
            "box_thresh": 0.6,
            "max_candidates": 1000,
            "unclip_ratio": 1.5,
            "min_size": 3
        })

    def preprocess(self, input_dicts, data_id, log_id):
        (_, input_dict), = input_dicts.items()
        data = base64.b64decode(input_dict["image"].encode('utf8'))
        self.raw_im = data
        data = np.fromstring(data, np.uint8)
        # Note: class variables(self.var) can only be used in process op mode
        im = cv2.imdecode(data, cv2.IMREAD_COLOR)
        self.ori_h, self.ori_w, _ = im.shape
        det_img = self.det_preprocess(im)
        _, self.new_h, self.new_w = det_img.shape
        return {
    
    "x": det_img[np.newaxis, :].copy()}, False, None, ""

    def postprocess(self, input_dicts, fetch_dict, data_id, log_id):
        det_out = list(fetch_dict.values())[0]
        ratio_list = [
            float(self.new_h) / self.ori_h, float(self.new_w) / self.ori_w
        ]
        dt_boxes_list = self.post_func(det_out, [ratio_list])
        dt_boxes = self.filter_func(dt_boxes_list[0], [self.ori_h, self.ori_w])
        out_dict = {
    
    "dt_boxes": dt_boxes, "image": self.raw_im}
        return out_dict, None, ""


class RecOp(Op):
    def init_op(self):
        self.ocr_reader = OCRReader(
            char_dict_path="./ppocr_keys_v1.txt")

        self.get_rotate_crop_image = GetRotateCropImage()
        self.sorted_boxes = SortedBoxes()

    def preprocess(self, input_dicts, data_id, log_id):
        (_, input_dict), = input_dicts.items()
        raw_im = input_dict["image"]
        data = np.frombuffer(raw_im, np.uint8)
        im = cv2.imdecode(data, cv2.IMREAD_COLOR)
        self.dt_list = input_dict["dt_boxes"]
        self.dt_list = self.sorted_boxes(self.dt_list)
        # deepcopy to save origin dt_boxes
        dt_boxes = copy.deepcopy(self.dt_list)
        feed_list = []
        img_list = []
        max_wh_ratio = 320/48.
        ## Many mini-batchs, the type of feed_data is list.
        max_batch_size = 6  # len(dt_boxes)

        # If max_batch_size is 0, skipping predict stage
        if max_batch_size == 0:
            return {
    
    }, True, None, ""
        boxes_size = len(dt_boxes)
        batch_size = boxes_size // max_batch_size
        rem = boxes_size % max_batch_size
        for bt_idx in range(0, batch_size + 1):
            imgs = None
            boxes_num_in_one_batch = 0
            if bt_idx == batch_size:
                if rem == 0:
                    continue
                else:
                    boxes_num_in_one_batch = rem
            elif bt_idx < batch_size:
                boxes_num_in_one_batch = max_batch_size
            else:
                _LOGGER.error("batch_size error, bt_idx={}, batch_size={}".
                              format(bt_idx, batch_size))
                break

            start = bt_idx * max_batch_size
            end = start + boxes_num_in_one_batch
            img_list = []
            for box_idx in range(start, end):
                boximg = self.get_rotate_crop_image(im, dt_boxes[box_idx])
                img_list.append(boximg)
                h, w = boximg.shape[0:2]
                wh_ratio = w * 1.0 / h
                max_wh_ratio = max(max_wh_ratio, wh_ratio)
            _, w, h = self.ocr_reader.resize_norm_img(img_list[0],
                                                      max_wh_ratio).shape

            imgs = np.zeros((boxes_num_in_one_batch, 3, w, h)).astype('float32')
            for id, img in enumerate(img_list):
                norm_img = self.ocr_reader.resize_norm_img(img, max_wh_ratio)
                imgs[id] = norm_img
            feed = {
    
    "x": imgs.copy()}
            feed_list.append(feed)
        return feed_list, False, None, ""

    def postprocess(self, input_dicts, fetch_data, data_id, log_id):
        rec_list = []
        dt_num = len(self.dt_list)
        if isinstance(fetch_data, dict):
            if len(fetch_data) > 0:
                rec_batch_res = self.ocr_reader.postprocess(
                    fetch_data, with_score=True)
                for res in rec_batch_res:
                    rec_list.append(res)
        elif isinstance(fetch_data, list):
            for one_batch in fetch_data:
                one_batch_res = self.ocr_reader.postprocess(
                    one_batch, with_score=True)
                for res in one_batch_res:
                    rec_list.append(res)
        result_list = []
        for i in range(dt_num):
            text = rec_list[i]
            dt_box = self.dt_list[i]
            if text[1] >= 0.5:
                result_list.append([text, dt_box.tolist()])
        res = {
    
    "result": str(result_list)}
        return res, None, ""


class OcrService(WebService):
    def get_pipeline_response(self, read_op):
        det_op = DetOp(name="det", input_ops=[read_op])
        rec_op = RecOp(name="rec", input_ops=[det_op])
        return rec_op


uci_service = OcrService(name="ocr")
uci_service.prepare_pipeline_config("config.yml")
uci_service.run_service()

1.5、ppocr_keys_v1字典下载地址

官方下载地址

1.6、安装包

paddle-serving-client==0.9.0
paddle-serving-app==0.9.0
paddle-serving-server-gpu==0.9.0.post101

1.7、ocr_reader.py

参考：官方下载地址

2、客户端部署

2.1、dockerfile编写

FROM registry.baidubce.com/paddlepaddle/serving:0.9.0-cuda10.1-cudnn7-devel

COPY . /deploy
WORKDIR /deploy

RUN pip config set global.index-url https://mirror.baidu.com/pypi/simple \
    && pip3.7 install --upgrade setuptools \
    && pip3.7 install --upgrade pip \
    && pip3.7 install -r requirements.txt

EXPOSE 9539
ENTRYPOINT ["gunicorn", "-c", "gunicorn_cfg.py", "main_api:app"]

2.2、制作镜像，启动容器指令

nvidia-docker build -t xx/ocr-python-client-qa:1.0.0.0630 .

nvidia-docker run -p 9539:9539 --name ocr-python-client-qa -d xx/ocr-python-client-qa:1.0.0.0630

nvidia-docker exec -it ocr-python-client-qa /bin/bash

2.3、pipeline_rpc_client.py

# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
try:
    from paddle_serving_server_gpu.pipeline import PipelineClient
except ImportError:
    from paddle_serving_server.pipeline import PipelineClient
import base64
import re
from config_env.config_envi import common


class ImageOCR(object):
    def __init__(self):
        self.client = PipelineClient()
        self.client.connect([common['model_ip_port_server']])
        self.index = ['姓名', '性别', '民族', '出生', '住址', '公民身份号码', '$']

    def cv2_to_base64(self, image):
        return base64.b64encode(image).decode('utf8')

    def ocr_image(self, img_path):
        with open(img_path, 'rb') as file:
            image_data = file.read()
        image = self.cv2_to_base64(image_data)
        ret = self.client.predict(feed_dict={
    
    "image": image}, fetch=["res"])
        res_new = eval(ret.value[0])
        res_list = []
        for txt in res_new:
            res_list.append(txt[0][0])
        res_list[0], res_list[1] = res_list[1], res_list[0]
        res_str = ''.join(res_list)
        res_dict = {
    
    }
        for i in range(len(self.index) - 1):
            re_res = re.search(r'%s(.*?)%s' % (self.index[i], self.index[i + 1]), res_str)
            # print(re_res)
            res_dict[self.index[i]] = re.sub(r'[\s。]', '', re_res.group(1)) if re_res else ''
        return res_dict


if __name__ == '__main__':
    image_ocr = ImageOCR()
    test_img_path_ = "./11.jpg"
    print(image_ocr.ocr_image(test_img_path_))

2.4、安装包

paddle-serving-client==0.9.0
paddle-serving-app==0.9.0
paddle-serving-server-gpu==0.9.0.post101
Flask-RESTful==0.3.9
Flask==1.1.4
flask_cors==3.0.10
prometheus-flask-exporter==0.18.7
gunicorn==20.1.0
gevent==21.12.0

其他参考，之前系列文章

一、OCR-docker部署最新版PaddleServing和PaddleOCR，服务端，客户端，dockerfile部署
二、OCR-docker部署最新版PaddleServing和PaddleOCR，服务端，客户端，dockerfile部署-C++部署
三、OCR-docker部署最新版PaddleServing和PaddleOCR，服务端，客户端，dockerfile部署-hubserving部署，GPU运行