add vlpr module (#1329)

modules/thirdparty/image/text_recognition/Vehicle_License_Plate_Recognition/README.md

+## 概述
+Vehicle_License_Plate_Recognition 是一个基于 CCPD 数据集训练的车牌识别模型，能够检测出图像中车牌位置并识别其中的车牌文字信息，大致的模型结构如下，分为检测车牌和文字识别两个模块：
+
+![](https://ai-studio-static-online.cdn.bcebos.com/35a3dab32ac948549de41afba7b51a5770d3f872d60b437d891f359a5cef8052)
+
+## API
+```python
+def plate_recognition(images)
+```
+车牌识别 API
+
+**参数**
+* images(str / ndarray / list(str) / list(ndarray))：待识别图像的路径或者图像的 Ndarray(RGB)
+
+**返回**
+* results(list(dict{'license', 'bbox'})): 识别到的车牌信息列表，包含车牌的位置坐标和车牌号码
+
+**代码示例**
+```python
+import paddlehub as hub
+
+# 加载模型
+model = hub.Module(name='Vehicle_License_Plate_Recognition')
+
+# 车牌识别
+result = model.plate_recognition("test.jpg")
+
+# 打印结果
+print(result)
+```
+    [{'license': '苏B92912', 'bbox': [[131.0, 251.0], [368.0, 253.0], [367.0, 338.0], [131.0, 336.0]]}]
+
+## 服务部署
+
+PaddleHub Serving 可以部署一个在线车牌识别服务。
+
+## 第一步：启动PaddleHub Serving
+
+运行启动命令：
+```shell
+$ hub serving start --modules Vehicle_License_Plate_Recognition
+```
+
+这样就完成了一个车牌识别的在线服务API的部署，默认端口号为8866。
+
+**NOTE:** 如使用GPU预测，则需要在启动服务之前，请设置CUDA\_VISIBLE\_DEVICES环境变量，否则不用设置。
+
+## 第二步：发送预测请求
+
+配置好服务端，以下数行代码即可实现发送预测请求，获取预测结果
+
+```python
+import requests
+import json
+import cv2
+import base64
+
+
+def cv2_to_base64(image):
+    data = cv2.imencode('.jpg', image)[1]
+    return base64.b64encode(data.tostring()).decode('utf8')
+
+
+# 发送HTTP请求
+data = {'images':[cv2_to_base64(cv2.imread("test.jpg"))]}
+headers = {"Content-type": "application/json"}
+url = "http://127.0.0.1:8866/predict/Vehicle_License_Plate_Recognition"
+r = requests.post(url=url, headers=headers, data=json.dumps(data))
+
+# 打印预测结果
+print(r.json()["results"])
+```
+    [{'bbox': [[260.0, 100.0], [546.0, 104.0], [544.0, 200.0], [259.0, 196.0]], 'license': '苏DS0000'}]
+
+## 查看代码
+https://github.com/jm12138/License_plate_recognition
+
+## 依赖
+paddlepaddle >= 2.0.0
+
+paddlehub >= 2.0.4
+
+paddleocr >= 2.0.2
\ No newline at end of file

modules/thirdparty/image/text_recognition/Vehicle_License_Plate_Recognition/module.py

+import os
+import cv2
+import base64
+import numpy as np
+import paddle.nn as nn
+from paddleocr import PaddleOCR
+from paddlehub.module.module import moduleinfo, serving
+
+
+@moduleinfo(
+    name="Vehicle_License_Plate_Recognition",
+    type="CV/text_recognition",
+    author="jm12138",
+    author_email="",
+    summary="Vehicle_License_Plate_Recognition",
+    version="1.0.0"
+)
+class Vehicle_License_Plate_Recognition(nn.Layer):
+    def __init__(self):
+        super(Vehicle_License_Plate_Recognition, self).__init__()
+        self.vlpr = PaddleOCR(
+            det_model_dir=os.path.join(self.directory, 'det_vlpr'),
+            rec_model_dir=os.path.join(
+                self.directory, 'ch_ppocr_server_v2.0_rec_infer')
+        )
+
+    @staticmethod
+    def base64_to_cv2(b64str):
+        data = base64.b64decode(b64str.encode('utf8'))
+        data = np.frombuffer(data, np.uint8)
+        data = cv2.imdecode(data, cv2.IMREAD_COLOR)
+        return data
+
+    def plate_recognition(self, images=None):
+        assert isinstance(images, (list, str, np.ndarray))
+        results = []
+
+        if isinstance(images, list):
+            for item in images:
+                for bbox, text in self.vlpr.ocr(item):
+                    results.append({
+                        'license': text[0],
+                        'bbox': bbox
+                    })
+
+        elif isinstance(images, (str, np.ndarray)):
+            for bbox, text in self.vlpr.ocr(images):
+                results.append({
+                    'license': text[0],
+                    'bbox': bbox
+                })
+
+        return results
+
+    @serving
+    def serving_method(self, images):
+        if isinstance(images, list):
+            images_decode = [self.base64_to_cv2(image) for image in images]
+        elif isinstance(images, str):
+            images_decode = self.base64_to_cv2(images)
+
+        return self.plate_recognition(images_decode)