Merge branch 'dygraph' of https://github.com/PaddlePaddle/PaddleOCR into lock_seed

PPOCRLabel/PPOCRLabel.py

@@ -27,7 +27,12 @@ import json
 import cv2
 
 
+
 __dir__ = os.path.dirname(os.path.abspath(__file__))
+
+import numpy as np
+
+
 sys.path.append(__dir__)
 sys.path.append(os.path.abspath(os.path.join(__dir__, '../..')))
 sys.path.append("..")
@@ -78,7 +83,7 @@ class WindowMixin(object):
             addActions(menu, actions)
         return menu
 
-    def toolbar(self, title, actions=None):  
+    def toolbar(self, title, actions=None):
         toolbar = ToolBar(title)
         toolbar.setObjectName(u'%sToolBar' % title)
         # toolbar.setOrientation(Qt.Vertical)
@@ -98,7 +103,7 @@ class MainWindow(QMainWindow, WindowMixin):
 
         # Load setting in the main thread
         self.settings = Settings()
-        self.settings.load()  
+        self.settings.load()
         settings = self.settings
         self.lang = lang
         # Load string bundle for i18n
@@ -159,7 +164,7 @@ class MainWindow(QMainWindow, WindowMixin):
         filelistLayout = QVBoxLayout()
         filelistLayout.setContentsMargins(0, 0, 0, 0)
         filelistLayout.addWidget(self.fileListWidget)
-        
+
         self.AutoRecognition = QToolButton()
         self.AutoRecognition.setToolButtonStyle(Qt.ToolButtonTextBesideIcon)
         self.AutoRecognition.setIcon(newIcon('Auto'))
@@ -176,7 +181,7 @@ class MainWindow(QMainWindow, WindowMixin):
         self.filedock.setObjectName(getStr('files'))
         self.filedock.setWidget(fileListContainer)
         self.addDockWidget(Qt.LeftDockWidgetArea, self.filedock)
-        
+
         ######## Right area ##########
         listLayout = QVBoxLayout()
         listLayout.setContentsMargins(0, 0, 0, 0)
@@ -250,7 +255,7 @@ class MainWindow(QMainWindow, WindowMixin):
         self.imgsplider.setMaximum(150)
         self.imgsplider.setSingleStep(1)
         self.imgsplider.setTickPosition(QSlider.TicksBelow)
-        self.imgsplider.setTickInterval(1) 
+        self.imgsplider.setTickInterval(1)
         op = QGraphicsOpacityEffect()
         op.setOpacity(0.2)
         self.imgsplider.setGraphicsEffect(op)
@@ -266,7 +271,9 @@ class MainWindow(QMainWindow, WindowMixin):
         self.zoomWidget = ZoomWidget()
         self.colorDialog = ColorDialog(parent=self)
         self.zoomWidgetValue = self.zoomWidget.value()
-        
+
+        self.msgBox = QMessageBox()
+
         ########## thumbnail #########
         hlayout = QHBoxLayout()
         m = (0, 0, 0, 0)
@@ -294,7 +301,7 @@ class MainWindow(QMainWindow, WindowMixin):
         self.nextButton.setStyleSheet('border: none;')
         self.nextButton.clicked.connect(self.openNextImg)
         self.nextButton.setShortcut('d')
-        
+
         hlayout.addWidget(self.preButton)
         hlayout.addWidget(self.iconlist)
         hlayout.addWidget(self.nextButton)
@@ -303,7 +310,7 @@ class MainWindow(QMainWindow, WindowMixin):
         iconListContainer = QWidget()
         iconListContainer.setLayout(hlayout)
         iconListContainer.setFixedHeight(100)
-        
+
         ########### Canvas ###########
         self.canvas = Canvas(parent=self)
         self.canvas.zoomRequest.connect(self.zoomRequest)
@@ -360,6 +367,9 @@ class MainWindow(QMainWindow, WindowMixin):
         opendir = action(getStr('openDir'), self.openDirDialog,
                          'Ctrl+u', 'open', getStr('openDir'))
 
+        open_dataset_dir = action(getStr('openDatasetDir'), self.openDatasetDirDialog,
+                         'Ctrl+p', 'open', getStr('openDatasetDir'), enabled=False)
+
         save = action(getStr('save'), self.saveFile,
                       'Ctrl+V', 'verify', getStr('saveDetail'), enabled=False)
 
@@ -439,7 +449,7 @@ class MainWindow(QMainWindow, WindowMixin):
         AutoRec = action(getStr('autoRecognition'), self.autoRecognition,
                       '', 'Auto', getStr('autoRecognition'), enabled=False)
 
-        reRec = action(getStr('reRecognition'), self.reRecognition, 
+        reRec = action(getStr('reRecognition'), self.reRecognition,
                       'Ctrl+Shift+R', 'reRec', getStr('reRecognition'), enabled=False)
 
         singleRere = action(getStr('singleRe'), self.singleRerecognition,
@@ -457,6 +467,12 @@ class MainWindow(QMainWindow, WindowMixin):
         undoLastPoint = action(getStr("undoLastPoint"), self.canvas.undoLastPoint,
                                'Ctrl+Z', "undo", getStr("undoLastPoint"), enabled=False)
 
+        rotateLeft = action(getStr("rotateLeft"), partial(self.rotateImgAction,1),
+                               'Ctrl+Alt+L', "rotateLeft", getStr("rotateLeft"), enabled=False)
+
+        rotateRight = action(getStr("rotateRight"), partial(self.rotateImgAction,-1),
+                               'Ctrl+Alt+R', "rotateRight", getStr("rotateRight"), enabled=False)
+
         undo = action(getStr("undo"), self.undoShapeEdit,
                       'Ctrl+Z', "undo", getStr("undo"), enabled=False)
 
@@ -520,13 +536,14 @@ class MainWindow(QMainWindow, WindowMixin):
                               zoom=zoom, zoomIn=zoomIn, zoomOut=zoomOut, zoomOrg=zoomOrg,
                               fitWindow=fitWindow, fitWidth=fitWidth,
                               zoomActions=zoomActions, saveLabel=saveLabel,
-                              undo=undo, undoLastPoint=undoLastPoint,
+                              undo=undo, undoLastPoint=undoLastPoint,open_dataset_dir=open_dataset_dir,
+                              rotateLeft=rotateLeft,rotateRight=rotateRight,
                               fileMenuActions=(
-                                  opendir, saveLabel,  resetAll, quit),
+                                  opendir,  open_dataset_dir, saveLabel,  resetAll, quit),
                               beginner=(), advanced=(),
                               editMenu=(createpoly, edit, copy, delete,singleRere,None, undo, undoLastPoint,
-                                        None, color1, self.drawSquaresOption),
-                              beginnerContext=(create, edit, copy, delete, singleRere),
+                                        None, rotateLeft, rotateRight, None, color1, self.drawSquaresOption),
+                              beginnerContext=(create, edit, copy, delete, singleRere, rotateLeft, rotateRight,),
                               advancedContext=(createMode, editMode, edit, copy,
                                                delete, shapeLineColor, shapeFillColor),
                               onLoadActive=(
@@ -564,7 +581,7 @@ class MainWindow(QMainWindow, WindowMixin):
         self.autoSaveOption.triggered.connect(self.autoSaveFunc)
 
         addActions(self.menus.file,
-                   (opendir, None, saveLabel, saveRec, self.autoSaveOption, None, resetAll, deleteImg, quit))
+                   (opendir, open_dataset_dir, None, saveLabel, saveRec, self.autoSaveOption, None, resetAll, deleteImg, quit))
 
         addActions(self.menus.help, (showKeys,showSteps, showInfo))
         addActions(self.menus.view, (
@@ -778,6 +795,38 @@ class MainWindow(QMainWindow, WindowMixin):
         self.actions.create.setEnabled(False)
         self.actions.undoLastPoint.setEnabled(True)
 
+    def rotateImg(self, filename, k, _value):
+
+        self.actions.rotateRight.setEnabled(_value)
+        pix = cv2.imread(filename)
+        pix = np.rot90(pix, k)
+        cv2.imwrite(filename, pix)
+        self.canvas.update()
+        self.loadFile(filename)
+
+    def rotateImgWarn(self):
+        if self.lang == 'ch':
+            self.msgBox.warning (self, "提示", "\n 该图片已经有标注框,旋转操作会打乱标注,建议清除标注框后旋转。")
+        else:
+            self.msgBox.warning (self, "Warn", "\n The picture already has a label box, and rotation will disrupt the label.\
+             It is recommended to clear the label box and rotate it.")
+
+    def rotateImgAction(self, k=1, _value=False):
+
+        filename = self.mImgList[self.currIndex]
+
+        if os.path.exists(filename):
+            if self.itemsToShapesbox:
+                self.rotateImgWarn()
+            else:
+                self.saveFile()
+                self.dirty = False
+                self.rotateImg(filename=filename, k=k, _value=True)
+        else:
+            self.rotateImgWarn()
+            self.actions.rotateRight.setEnabled(False)
+            self.actions.rotateLeft.setEnabled(False)
+
     def toggleDrawingSensitive(self, drawing=True):
         """In the middle of drawing, toggling between modes should be disabled."""
         self.actions.editMode.setEnabled(not drawing)
@@ -885,7 +934,12 @@ class MainWindow(QMainWindow, WindowMixin):
             self.updateComboBox()
 
     def updateBoxlist(self):
-        for shape in self.canvas.selectedShapes+[self.canvas.hShape]:
+        self.canvas.selectedShapes_hShape = []
+        if self.canvas.hShape != None:
+            self.canvas.selectedShapes_hShape = self.canvas.selectedShapes + [self.canvas.hShape]
+        else:
+            self.canvas.selectedShapes_hShape = self.canvas.selectedShapes
+        for shape in self.canvas.selectedShapes_hShape:
             item = self.shapesToItemsbox[shape]  # listitem
             text = [(int(p.x()), int(p.y())) for p in shape.points]
             item.setText(str(text))
@@ -1274,7 +1328,7 @@ class MainWindow(QMainWindow, WindowMixin):
                         titem = self.iconlist.item(i)
                         titem.setSelected(True)
                         self.iconlist.scrollToItem(titem)
-                        break 
+                        break
             else:
                 self.fileListWidget.clear()
                 self.mImgList.clear()
@@ -1282,7 +1336,7 @@ class MainWindow(QMainWindow, WindowMixin):
 
         # if unicodeFilePath and self.iconList.count() > 0:
         #     if unicodeFilePath in self.mImgList:
-                
+
         if unicodeFilePath and os.path.exists(unicodeFilePath):
             self.canvas.verified = False
 
@@ -1313,7 +1367,7 @@ class MainWindow(QMainWindow, WindowMixin):
             self.addRecentFile(self.filePath)
             self.toggleActions(True)
             self.showBoundingBoxFromPPlabel(filePath)
-            
+
             self.setWindowTitle(__appname__ + ' ' + filePath)
 
             # Default : select last item if there is at least one item
@@ -1325,7 +1379,7 @@ class MainWindow(QMainWindow, WindowMixin):
             return True
         return False
 
-    
+
     def showBoundingBoxFromPPlabel(self, filePath):
         imgidx = self.getImglabelidx(filePath)
         if imgidx not in self.PPlabel.keys():
@@ -1418,6 +1472,7 @@ class MainWindow(QMainWindow, WindowMixin):
 
     def loadRecent(self, filename):
         if self.mayContinue():
+            print(filename,"======")
             self.loadFile(filename)
 
     def scanAllImages(self, folderPath):
@@ -1453,6 +1508,23 @@ class MainWindow(QMainWindow, WindowMixin):
         self.lastOpenDir = targetDirPath
         self.importDirImages(targetDirPath)
 
+    def openDatasetDirDialog(self,):
+        if self.lastOpenDir and os.path.exists(self.lastOpenDir):
+            if platform.system() == 'Windows':
+                os.startfile(self.lastOpenDir)
+            else:
+                os.system('open ' + os.path.normpath(self.lastOpenDir))
+            defaultOpenDirPath = self.lastOpenDir
+
+        else:
+            if self.lang == 'ch':
+                self.msgBox.warning(self, "提示", "\n 原文件夹已不存在,请从新选择数据集路径!")
+            else:
+                self.msgBox.warning(self, "Warn", "\n The original folder no longer exists, please choose the data set path again!")
+
+            self.actions.open_dataset_dir.setEnabled(False)
+            defaultOpenDirPath = os.path.dirname(self.filePath) if self.filePath else '.'
+
     def importDirImages(self, dirpath, isDelete = False):
         if not self.mayContinue() or not dirpath:
             return
@@ -1500,6 +1572,10 @@ class MainWindow(QMainWindow, WindowMixin):
         self.reRecogButton.setEnabled(True)
         self.actions.AutoRec.setEnabled(True)
         self.actions.reRec.setEnabled(True)
+        self.actions.open_dataset_dir.setEnabled(True)
+        self.actions.rotateLeft.setEnabled(True)
+        self.actions.rotateRight.setEnabled(True)
+
 
 
     def openPrevImg(self, _value=False):
@@ -1508,7 +1584,7 @@ class MainWindow(QMainWindow, WindowMixin):
 
         if self.filePath is None:
             return
-        
+
         currIndex = self.mImgList.index(self.filePath)
         self.mImgList5 = self.mImgList[:5]
         if currIndex - 1 >= 0:
@@ -1538,7 +1614,7 @@ class MainWindow(QMainWindow, WindowMixin):
         if filename:
             print('file name in openNext is ',filename)
             self.loadFile(filename)
-        
+
     def updateFileListIcon(self, filename):
         pass
 
@@ -1650,7 +1726,7 @@ class MainWindow(QMainWindow, WindowMixin):
         proc.startDetached(os.path.abspath(__file__))
 
     def mayContinue(self):  #
-        if not self.dirty:                                    
+        if not self.dirty:
             return True
         else:
             discardChanges = self.discardChangesDialog()
@@ -2077,7 +2153,7 @@ def main():
 
 
 if __name__ == '__main__':
-        
+
     resource_file = './libs/resources.py'
     if not os.path.exists(resource_file):
         output = os.system('pyrcc5 -o libs/resources.py resources.qrc')

PPOCRLabel/README.md

@@ -8,9 +8,12 @@ PPOCRLabel is a semi-automatic graphic annotation tool suitable for OCR field, w
 
 ### Recent Update
 
+- 2021.8.11：
+  - New functions: Open the dataset folder, image rotation (Note: Please delete the label box before rotating the image) (by [Wei-JL](https://github.com/Wei-JL))
+  - Added shortcut key description (Help-Shortcut Key), repaired the direction shortcut key movement function under batch processing (by [d2623587501](https://github.com/d2623587501))
 - 2021.2.5: New batch processing and undo functions (by [Evezerest](https://github.com/Evezerest)):
-  - Batch processing function: Press and hold the Ctrl key to select the box, you can move, copy, and delete in batches.
-  - Undo function: In the process of drawing a four-point label box or after editing the box, press Ctrl+Z to undo the previous operation.
+  - **Batch processing function**: Press and hold the Ctrl key to select the box, you can move, copy, and delete in batches.
+  - **Undo function**: In the process of drawing a four-point label box or after editing the box, press Ctrl+Z to undo the previous operation.
   - Fix image rotation and size problems, optimize the process of editing the mark frame (by [ninetailskim](https://github.com/ninetailskim)、 [edencfc](https://github.com/edencfc)).
 - 2021.1.11: Optimize the labeling experience (by [edencfc](https://github.com/edencfc)),
   - Users can choose whether to pop up the label input dialog after drawing the detection box in "View - Pop-up Label Input Dialog".
@@ -23,15 +26,51 @@ PPOCRLabel is a semi-automatic graphic annotation tool suitable for OCR field, w
 
 ## Installation
 
-### 1. Install PaddleOCR
+### 1. Environment Preparation
 
-PaddleOCR models has been built in PPOCRLabel, please refer to [PaddleOCR installation document](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/doc/doc_ch/installation.md) to prepare PaddleOCR and make sure it works.
+#### **Install PaddlePaddle 2.0**
+
+```bash
+pip3 install --upgrade pip
+
+# If you have cuda9 or cuda10 installed on your machine, please run the following command to install
+python3 -m pip install paddlepaddle-gpu==2.0.0 -i https://mirror.baidu.com/pypi/simple
+
+# If you only have cpu on your machine, please run the following command to install
+python3 -m pip install paddlepaddle==2.0.0 -i https://mirror.baidu.com/pypi/simple
+```
+
+For more software version requirements, please refer to the instructions in [Installation Document](https://www.paddlepaddle.org.cn/install/quick) for operation.
+
+#### **Install PaddleOCR**
+
+```bash
+# Recommend
+git clone https://github.com/PaddlePaddle/PaddleOCR
+
+# If you cannot pull successfully due to network problems, you can also choose to use the code hosting on the cloud:
+
+git clone https://gitee.com/paddlepaddle/PaddleOCR
+
+# Note: The cloud-hosting code may not be able to synchronize the update with this GitHub project in real time. There might be a delay of 3-5 days. Please give priority to the recommended method.
+```
+
+#### **Install Third-party Libraries**
+
+```bash
+cd PaddleOCR
+pip3 install -r requirements.txt
+```
+
+If you getting this error `OSError: [WinError 126] The specified module could not be found` when you install shapely on windows. Please try to download Shapely whl file using http://www.lfd.uci.edu/~gohlke/pythonlibs/#shapely.
+
+Reference: [Solve shapely installation on windows](https://stackoverflow.com/questions/44398265/install-shapely-oserror-winerror-126-the-specified-module-could-not-be-found)
 
 ### 2. Install PPOCRLabel
 
 #### Windows
 
-```
+```bash
 pip install pyqt5
 cd ./PPOCRLabel # Change the directory to the PPOCRLabel folder
 python PPOCRLabel.py
@@ -39,15 +78,15 @@ python PPOCRLabel.py
 
 #### Ubuntu Linux
 
-```
+```bash
 pip3 install pyqt5
 pip3 install trash-cli
 cd ./PPOCRLabel # Change the directory to the PPOCRLabel folder
 python3 PPOCRLabel.py
 ```
 
-#### macOS
-```
+#### MacOS
+```bash
 pip3 install pyqt5
 pip3 uninstall opencv-python # Uninstall opencv manually as it conflicts with pyqt
 pip3 install opencv-contrib-python-headless==4.2.0.32 # Install the headless version of opencv
@@ -77,11 +116,11 @@ python3 PPOCRLabel.py
 
 7. Double click the result in 'recognition result' list to manually change inaccurate recognition results.
 
-8. Click "Check", the image status will switch to "√",then the program automatically jump to the next.
+8. **Click "Check", the image status will switch to "√",then the program automatically jump to the next.**
 
 9. Click "Delete Image" and the image will be deleted to the recycle bin.
 
-10. Labeling result: the user can save manually through the menu "File - Save Label", while the program will also save automatically if "File - Auto Save Label Mode" is selected. The manually checked label will be stored in *Label.txt* under the opened picture folder. Click "PaddleOCR"-"Save Recognition Results" in the menu bar, the recognition training data of such pictures will be saved in the *crop_img* folder, and the recognition label will be saved in *rec_gt.txt*<sup>[4]</sup>.
+10. Labeling result: the user can export the label result manually through the menu "File - Export Label", while the program will also export automatically if "File - Auto export Label Mode" is selected. The manually checked label will be stored in *Label.txt* under the opened picture folder. Click "File"-"Export Recognition Results" in the menu bar, the recognition training data of such pictures will be saved in the *crop_img* folder, and the recognition label will be saved in *rec_gt.txt*<sup>[4]</sup>.
 
 ### Note
 
@@ -95,10 +134,10 @@ python3 PPOCRLabel.py
 
 |   File name   |                         Description                          |
 | :-----------: | :----------------------------------------------------------: |
-|   Label.txt   | The detection label file can be directly used for PPOCR detection model training. After the user saves 5 label results, the file will be automatically saved. It will also be written when the user closes the application or changes the file folder. |
+|   Label.txt   | The detection label file can be directly used for PPOCR detection model training. After the user saves 5 label results, the file will be automatically exported. It will also be written when the user closes the application or changes the file folder. |
 | fileState.txt | The picture status file save the image in the current folder that has been manually confirmed by the user. |
 |  Cache.cach   |    Cache files to save the results of model recognition.     |
-|  rec_gt.txt   | The recognition label file, which can be directly used for PPOCR identification model training, is generated after the user clicks on the menu bar "File"-"Save recognition result". |
+|  rec_gt.txt   | The recognition label file, which can be directly used for PPOCR identification model training, is generated after the user clicks on the menu bar "File"-"Export recognition result". |
 |   crop_img    | The recognition data, generated at the same time with *rec_gt.txt* |
 
 ## Explanation
@@ -132,16 +171,16 @@ python3 PPOCRLabel.py
 
 - Custom model: The model trained by users can be replaced by modifying PPOCRLabel.py in [PaddleOCR class instantiation](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/PPOCRLabel/PPOCRLabel.py#L110) referring [Custom Model Code](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/doc/doc_en/whl_en.md#use-custom-model)
 
-### Save
+### Export Label Result
 
-PPOCRLabel supports three ways to save Label.txt
+PPOCRLabel supports three ways to export Label.txt
 
-- Automatically save: After selecting "File - Auto Save Label Mode", the program will automatically write the annotations into Label.txt every time the user confirms an image. If this option is not turned on, it will be automatically saved after detecting that the user has manually checked 5 images.
-- Manual save: Click "File-Save Marking Results" to manually save the label.
-- Close application save
+- Automatically export: After selecting "File - Auto Export Label Mode", the program will automatically write the annotations into Label.txt every time the user confirms an image. If this option is not turned on, it will be automatically exported after detecting that the user has manually checked 5 images.
+- Manual export: Click "File-Export Marking Results" to manually export the label.
+- Close application export
 
 
-### Export partial recognition results
+### Export Partial Recognition Results
 
 For some data that are difficult to recognize, the recognition results will not be exported by **unchecking** the corresponding tags in the recognition results checkbox.
 

PPOCRLabel/README_ch.md

@@ -8,9 +8,12 @@ PPOCRLabel是一款适用于OCR领域的半自动化图形标注工具，内置P
 
 #### 近期更新
 
+- 2021.8.11：
+  - 新增功能：打开数据所在文件夹、图像旋转（注意：旋转前的图片上不能存在标记框）（by [Wei-JL](https://github.com/Wei-JL)）
+  - 新增快捷键说明（帮助-快捷键）、修复批处理下的方向快捷键移动功能（by [d2623587501](https://github.com/d2623587501)）
 - 2021.2.5：新增批处理与撤销功能（by [Evezerest](https://github.com/Evezerest))
-  - 批处理功能：按住Ctrl键选择标记框后可批量移动、复制、删除。
-  - 撤销功能：在绘制四点标注框过程中或对框进行编辑操作后，按下Ctrl+Z可撤销上一部操作。
+  - **批处理功能**：按住Ctrl键选择标记框后可批量移动、复制、删除、重新识别。
+  - **撤销功能**：在绘制四点标注框过程中或对框进行编辑操作后，按下Ctrl+Z可撤销上一部操作。
   - 修复图像旋转和尺寸问题、优化编辑标记框过程（by [ninetailskim](https://github.com/ninetailskim)、 [edencfc](https://github.com/edencfc)）
 - 2021.1.11：优化标注体验（by [edencfc](https://github.com/edencfc)）：
   - 用户可在“视图 - 弹出标记输入框”选择在画完检测框后标记输入框是否弹出。
@@ -27,13 +30,48 @@ PPOCRLabel是一款适用于OCR领域的半自动化图形标注工具，内置P
 
 ## 安装
 
-### 1. 安装PaddleOCR
-PPOCRLabel内置PaddleOCR模型，故请参考[PaddleOCR安装文档](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/doc/doc_ch/installation.md)准备好PaddleOCR，并确保PaddleOCR安装成功。
+### 1. 环境搭建
+#### 安装PaddlePaddle
+
+```bash
+pip3 install --upgrade pip
+
+如果您的机器安装的是CUDA9或CUDA10，请运行以下命令安装
+python3 -m pip install paddlepaddle-gpu==2.0.0 -i https://mirror.baidu.com/pypi/simple
+
+如果您的机器是CPU，请运行以下命令安装
+
+python3 -m pip install paddlepaddle==2.0.0 -i https://mirror.baidu.com/pypi/simple
+```
+
+更多的版本需求，请参照[安装文档](https://www.paddlepaddle.org.cn/install/quick)中的说明进行操作。
+
+#### **安装PaddleOCR**
+
+```bash
+【推荐】git clone https://github.com/PaddlePaddle/PaddleOCR
+
+如果因为网络问题无法pull成功，也可选择使用码云上的托管：
+
+git clone https://gitee.com/paddlepaddle/PaddleOCR
+
+注：码云托管代码可能无法实时同步本github项目更新，存在3~5天延时，请优先使用推荐方式。
+```
+
+#### 安装第三方库
+
+```bash
+cd PaddleOCR
+pip3 install -r requirements.txt
+```
+
+注意，windows环境下，建议从[这里](https://www.lfd.uci.edu/~gohlke/pythonlibs/#shapely)下载shapely安装包完成安装， 直接通过pip安装的shapely库可能出现`[winRrror 126] 找不到指定模块的问题`。
 
 ### 2. 安装PPOCRLabel
+
 #### Windows
 
-```
+```bash
 pip install pyqt5
 cd ./PPOCRLabel # 将目录切换到PPOCRLabel文件夹下
 python PPOCRLabel.py --lang ch
@@ -41,15 +79,15 @@ python PPOCRLabel.py --lang ch
 
 #### Ubuntu Linux
 
-```
+```bash
 pip3 install pyqt5
 pip3 install trash-cli
 cd ./PPOCRLabel # 将目录切换到PPOCRLabel文件夹下
 python3 PPOCRLabel.py --lang ch
 ```
 
-#### macOS
-```
+#### MacOS
+```bash
 pip3 install pyqt5
 pip3 uninstall opencv-python # 由于mac版本的opencv与pyqt有冲突，需先手动卸载opencv
 pip3 install opencv-contrib-python-headless==4.2.0.32 # 安装headless版本的open-cv
@@ -57,6 +95,8 @@ cd ./PPOCRLabel # 将目录切换到PPOCRLabel文件夹下
 python3 PPOCRLabel.py --lang ch
 ```
 
+
+
 ## 使用
 
 ### 操作步骤
@@ -68,9 +108,9 @@ python3 PPOCRLabel.py --lang ch
 5. 标记框绘制完成后，用户点击 “确认”，检测框会先被预分配一个 “待识别” 标签。
 6. 重新识别：将图片中的所有检测画绘制/调整完成后，点击 “重新识别”，PPOCR模型会对当前图片中的**所有检测框**重新识别<sup>[3]</sup>。
 7. 内容更改：双击识别结果，对不准确的识别结果进行手动更改。
-8. **确认标记**：点击 “确认”，图片状态切换为 “√”，跳转至下一张。
+8. **确认标记：点击 “确认”，图片状态切换为 “√”，跳转至下一张。**
 9. 删除：点击 “删除图像”，图片将会被删除至回收站。
-10. 保存结果：用户可以通过菜单中“文件-保存标记结果”手动保存，同时也可以点击“文件 - 自动保存标记结果”开启自动保存。手动确认过的标记将会被存放在所打开图片文件夹下的*Label.txt*中。在菜单栏点击 “文件” - "保存识别结果"后，会将此类图片的识别训练数据保存在*crop_img*文件夹下，识别标签保存在*rec_gt.txt*中<sup>[4]</sup>。
+10. 导出结果：用户可以通过菜单中“文件-导出标记结果”手动导出，同时也可以点击“文件 - 自动导出标记结果”开启自动导出。手动确认过的标记将会被存放在所打开图片文件夹下的*Label.txt*中。在菜单栏点击 “文件” - "导出识别结果"后，会将此类图片的识别训练数据保存在*crop_img*文件夹下，识别标签保存在*rec_gt.txt*中<sup>[4]</sup>。
 
 ### 注意
 
@@ -84,10 +124,10 @@ python3 PPOCRLabel.py --lang ch
 
 |    文件名     |                             说明                             |
 | :-----------: | :----------------------------------------------------------: |
-|   Label.txt   | 检测标签，可直接用于PPOCR检测模型训练。用户每保存5张检测结果后，程序会进行自动写入。当用户关闭应用程序或切换文件路径后同样会进行写入。 |
+|   Label.txt   | 检测标签，可直接用于PPOCR检测模型训练。用户每确认5张检测结果后，程序会进行自动写入。当用户关闭应用程序或切换文件路径后同样会进行写入。 |
 | fileState.txt | 图片状态标记文件，保存当前文件夹下已经被用户手动确认过的图片名称。 |
 |  Cache.cach   |              缓存文件，保存模型自动识别的结果。              |
-|  rec_gt.txt   | 识别标签。可直接用于PPOCR识别模型训练。需用户手动点击菜单栏“文件” - "保存识别结果"后产生。 |
+|  rec_gt.txt   | 识别标签。可直接用于PPOCR识别模型训练。需用户手动点击菜单栏“文件” - "导出识别结果"后产生。 |
 |   crop_img    |   识别数据。按照检测框切割后的图片。与rec_gt.txt同时产生。   |
 
 ## 说明
@@ -120,19 +160,19 @@ python3 PPOCRLabel.py --lang ch
 
  - 自定义模型：用户可根据[自定义模型代码使用](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/doc/doc_ch/whl.md#%E8%87%AA%E5%AE%9A%E4%B9%89%E6%A8%A1%E5%9E%8B)，通过修改PPOCRLabel.py中针对[PaddleOCR类的实例化](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/PPOCRLabel/PPOCRLabel.py#L110)替换成自己训练的模型。
 
-### 保存方式
+### 导出标记结果
 
-PPOCRLabel支持三种保存方式：
+PPOCRLabel支持三种导出方式：
 
-- 自动保存：点击“文件 - 自动保存标记结果”后，用户每确认过一张图片，程序自动将标记结果写入Label.txt中。若未开启此选项，则检测到用户手动确认过5张图片后进行自动保存。
-- 手动保存：点击“文件 - 保存标记结果”手动保存标记。
-- 关闭应用程序保存
+- 自动导出：点击“文件 - 自动导出标记结果”后，用户每确认过一张图片，程序自动将标记结果写入Label.txt中。若未开启此选项，则检测到用户手动确认过5张图片后进行自动导出。
+- 手动导出：点击“文件 - 导出标记结果”手动导出标记。
+- 关闭应用程序导出
 
 ### 导出部分识别结果
 
 针对部分难以识别的数据，通过在识别结果的复选框中**取消勾选**相应的标记，其识别结果不会被导出。
 
-*注意：识别结果中的复选框状态仍需用户手动点击保存后才能保留*
+*注意：识别结果中的复选框状态仍需用户手动点击确认后才能保留*
 
 ### 错误提示
 - 如果同时使用whl包安装了paddleocr，其优先级大于通过paddleocr.py调用PaddleOCR类，whl包未更新时会导致程序异常。

PPOCRLabel/libs/canvas.py

@@ -23,6 +23,7 @@ except ImportError:
 
 from libs.shape import Shape
 from libs.utils import distance
+import copy
 
 CURSOR_DEFAULT = Qt.ArrowCursor
 CURSOR_POINT = Qt.PointingHandCursor
@@ -81,6 +82,7 @@ class Canvas(QWidget):
         self.fourpoint = True # ADD
         self.pointnum = 0
         self.movingShape = False
+        self.selectCountShape = False
 
         #initialisation for panning
         self.pan_initial_pos = QPoint()
@@ -702,6 +704,10 @@ class Canvas(QWidget):
 
     def keyPressEvent(self, ev):
         key = ev.key()
+        shapesBackup = []
+        shapesBackup = copy.deepcopy(self.shapes)
+        self.shapesBackups.pop()
+        self.shapesBackups.append(shapesBackup)
         if key == Qt.Key_Escape and self.current:
             print('ESC press')
             self.current = None
@@ -709,41 +715,48 @@ class Canvas(QWidget):
             self.update()
         elif key == Qt.Key_Return and self.canCloseShape():
             self.finalise()
-        elif key == Qt.Key_Left and self.selectedShape:
+        elif key == Qt.Key_Left and self.selectedShapes:
              self.moveOnePixel('Left')
-        elif key == Qt.Key_Right and self.selectedShape:
+        elif key == Qt.Key_Right and self.selectedShapes:
              self.moveOnePixel('Right')
-        elif key == Qt.Key_Up and self.selectedShape:
+        elif key == Qt.Key_Up and self.selectedShapes:
              self.moveOnePixel('Up')
-        elif key == Qt.Key_Down and self.selectedShape:
+        elif key == Qt.Key_Down and self.selectedShapes:
              self.moveOnePixel('Down')
 
     def moveOnePixel(self, direction):
         # print(self.selectedShape.points)
-        if direction == 'Left' and not self.moveOutOfBound(QPointF(-1.0, 0)):
-            # print("move Left one pixel")
-            self.selectedShape.points[0] += QPointF(-1.0, 0)
-            self.selectedShape.points[1] += QPointF(-1.0, 0)
-            self.selectedShape.points[2] += QPointF(-1.0, 0)
-            self.selectedShape.points[3] += QPointF(-1.0, 0)
-        elif direction == 'Right' and not self.moveOutOfBound(QPointF(1.0, 0)):
-            # print("move Right one pixel")
-            self.selectedShape.points[0] += QPointF(1.0, 0)
-            self.selectedShape.points[1] += QPointF(1.0, 0)
-            self.selectedShape.points[2] += QPointF(1.0, 0)
-            self.selectedShape.points[3] += QPointF(1.0, 0)
-        elif direction == 'Up' and not self.moveOutOfBound(QPointF(0, -1.0)):
-            # print("move Up one pixel")
-            self.selectedShape.points[0] += QPointF(0, -1.0)
-            self.selectedShape.points[1] += QPointF(0, -1.0)
-            self.selectedShape.points[2] += QPointF(0, -1.0)
-            self.selectedShape.points[3] += QPointF(0, -1.0)
-        elif direction == 'Down' and not self.moveOutOfBound(QPointF(0, 1.0)):
-            # print("move Down one pixel")
-            self.selectedShape.points[0] += QPointF(0, 1.0)
-            self.selectedShape.points[1] += QPointF(0, 1.0)
-            self.selectedShape.points[2] += QPointF(0, 1.0)
-            self.selectedShape.points[3] += QPointF(0, 1.0)
+        self.selectCount = len(self.selectedShapes)
+        self.selectCountShape = True
+        for i in range(len(self.selectedShapes)):
+            self.selectedShape = self.selectedShapes[i]
+            if direction == 'Left' and not self.moveOutOfBound(QPointF(-1.0, 0)):
+                # print("move Left one pixel")
+                self.selectedShape.points[0] += QPointF(-1.0, 0)
+                self.selectedShape.points[1] += QPointF(-1.0, 0)
+                self.selectedShape.points[2] += QPointF(-1.0, 0)
+                self.selectedShape.points[3] += QPointF(-1.0, 0)
+            elif direction == 'Right' and not self.moveOutOfBound(QPointF(1.0, 0)):
+                # print("move Right one pixel")
+                self.selectedShape.points[0] += QPointF(1.0, 0)
+                self.selectedShape.points[1] += QPointF(1.0, 0)
+                self.selectedShape.points[2] += QPointF(1.0, 0)
+                self.selectedShape.points[3] += QPointF(1.0, 0)
+            elif direction == 'Up' and not self.moveOutOfBound(QPointF(0, -1.0)):
+                # print("move Up one pixel")
+                self.selectedShape.points[0] += QPointF(0, -1.0)
+                self.selectedShape.points[1] += QPointF(0, -1.0)
+                self.selectedShape.points[2] += QPointF(0, -1.0)
+                self.selectedShape.points[3] += QPointF(0, -1.0)
+            elif direction == 'Down' and not self.moveOutOfBound(QPointF(0, 1.0)):
+                # print("move Down one pixel")
+                self.selectedShape.points[0] += QPointF(0, 1.0)
+                self.selectedShape.points[1] += QPointF(0, 1.0)
+                self.selectedShape.points[2] += QPointF(0, 1.0)
+                self.selectedShape.points[3] += QPointF(0, 1.0)
+        shapesBackup = []
+        shapesBackup = copy.deepcopy(self.shapes)
+        self.shapesBackups.append(shapesBackup)
         self.shapeMoved.emit()
         self.repaint()
 
@@ -840,6 +853,7 @@ class Canvas(QWidget):
     def restoreShape(self):
         if not self.isShapeRestorable:
             return
+
         self.shapesBackups.pop()  # latest
         shapesBackup = self.shapesBackups.pop()
         self.shapes = shapesBackup

PPOCRLabel/resources.qrc

@@ -18,6 +18,8 @@
 <file alias="quit">resources/icons/quit.png</file>
 <file alias="copy">resources/icons/copy.png</file>
 <file alias="edit">resources/icons/edit.png</file>
+<file alias="rotateLeft">resources/icons/rotateLeft.png</file>
+<file alias="rotateRight">resources/icons/rotateRight.png</file>
 <file alias="open">resources/icons/open.png</file>
 <file alias="save">resources/icons/save.png</file>
 <file alias="format_voc">resources/icons/format_voc.png</file>

PPOCRLabel/resources/strings/strings-zh-CN.properties

@@ -31,6 +31,7 @@ save=确认
 saveAs=另存为
 fitWinDetail=缩放到当前窗口大小
 openDir=打开目录
+openDatasetDir=打开数据集路径
 copyPrevBounding=复制当前图像中的上一个边界框
 showHide=显示/隐藏标签
 changeSaveFormat=更改存储格式
@@ -85,7 +86,9 @@ detectionBoxposition=检测框位置
 recognitionResult=识别结果
 creatPolygon=四点标注
 drawSquares=正方形标注
-saveRec=保存识别结果
+rotateLeft=图片左旋转90度
+rotateRight=图片右旋转90度
+saveRec=导出识别结果
 tempLabel=待识别
 nullLabel=无法识别
 steps=操作步骤
@@ -96,9 +99,9 @@ ok=确认
 autolabeling=自动标注中
 hideBox=隐藏所有标注
 showBox=显示所有标注
-saveLabel=保存标记结果
+saveLabel=导出标记结果
 singleRe=重识别此区块
 labelDialogOption=弹出标记输入框
 undo=撤销
 undoLastPoint=撤销上个点
-autoSaveMode=自动保存标记结果
\ No newline at end of file
+autoSaveMode=自动导出标记结果
\ No newline at end of file

PPOCRLabel/resources/strings/strings.properties

@@ -3,6 +3,7 @@ openFileDetail=Open image or label file
 quit=Quit
 quitApp=Quit application
 openDir=Open Dir
+openDatasetDir=Open DatasetDir
 copyPrevBounding=Copy previous Bounding Boxes in the current image 
 changeSavedAnnotationDir=Change default saved Annotation dir
 openAnnotation=Open Annotation
@@ -84,8 +85,10 @@ iconList=Icon List
 detectionBoxposition=Detection box position
 recognitionResult=Recognition result
 creatPolygon=Create Quadrilateral
+rotateLeft=Left turn 90 degrees
+rotateRight=Right turn 90 degrees
 drawSquares=Draw Squares
-saveRec=Save Recognition Result
+saveRec=Export Recognition Result
 tempLabel=TEMPORARY
 nullLabel=NULL
 steps=Steps
@@ -96,9 +99,9 @@ ok=OK
 autolabeling=Automatic Labeling
 hideBox=Hide All Box
 showBox=Show All Box
-saveLabel=Save Label
+saveLabel=Export Label
 singleRe=Re-recognition RectBox
 labelDialogOption=Pop-up Label Input Dialog
 undo=Undo
 undoLastPoint=Undo Last Point
-autoSaveMode=Auto Save Label Mode
\ No newline at end of file
+autoSaveMode=Auto Export Label Mode
\ No newline at end of file

configs/rec/rec_mtb_nrtr.yml

+Global:
+  use_gpu: True
+  epoch_num: 21
+  log_smooth_window: 20
+  print_batch_step: 10
+  save_model_dir: ./output/rec/nrtr/
+  save_epoch_step: 1
+  # evaluation is run every 2000 iterations
+  eval_batch_step: [0, 2000]
+  cal_metric_during_train: True
+  pretrained_model:
+  checkpoints: 
+  save_inference_dir:
+  use_visualdl: False
+  infer_img: doc/imgs_words_en/word_10.png
+  # for data or label process
+  character_dict_path: 
+  character_type: EN_symbol
+  max_text_length: 25
+  infer_mode: False
+  use_space_char: True
+  save_res_path: ./output/rec/predicts_nrtr.txt
+
+Optimizer:
+  name: Adam
+  beta1: 0.9
+  beta2: 0.99
+  clip_norm: 5.0
+  lr:
+    name: Cosine
+    learning_rate: 0.0005
+    warmup_epoch: 2
+  regularizer:
+    name: 'L2'
+    factor: 0.
+
+Architecture:
+  model_type: rec
+  algorithm: NRTR
+  in_channels: 1
+  Transform:
+  Backbone:
+    name: MTB
+    cnn_num: 2
+  Head:
+    name: Transformer
+    d_model: 512
+    num_encoder_layers: 6
+    beam_size: 10 # When Beam size is greater than 0, it means to use beam search when evaluation.
+    
+
+Loss:
+  name: NRTRLoss
+  smoothing: True
+
+PostProcess:
+  name: NRTRLabelDecode
+
+Metric:
+  name: RecMetric
+  main_indicator: acc
+
+Train:
+  dataset:
+    name: LMDBDataSet
+    data_dir: ./train_data/data_lmdb_release/training/
+    transforms:
+      - NRTRDecodeImage: # load image
+          img_mode: BGR
+          channel_first: False
+      - NRTRLabelEncode: # Class handling label
+      - NRTRRecResizeImg:
+          image_shape: [100, 32]
+          resize_type: PIL # PIL or OpenCV
+      - KeepKeys:
+          keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order
+  loader:
+    shuffle: True
+    batch_size_per_card: 512
+    drop_last: True
+    num_workers: 8
+
+Eval:
+  dataset:
+    name: LMDBDataSet
+    data_dir: ./train_data/data_lmdb_release/evaluation/
+    transforms:
+      - NRTRDecodeImage: # load image
+          img_mode: BGR
+          channel_first: False
+      - NRTRLabelEncode: # Class handling label
+      - NRTRRecResizeImg:
+          image_shape: [100, 32]
+          resize_type: PIL # PIL or OpenCV
+      - KeepKeys:
+          keep_keys: ['image', 'label', 'length'] # dataloader will return list in this order
+  loader:
+    shuffle: False
+    drop_last: False
+    batch_size_per_card: 256
+    num_workers: 1
+    use_shared_memory: False

deploy/cpp_infer/CMakeLists.txt

-project(ocr_system CXX C)
+project(ppocr CXX C)
 
 option(WITH_MKL        "Compile demo with MKL/OpenBlas support, default use MKL."       ON)
 option(WITH_GPU        "Compile demo with GPU/CPU, default use CPU."                    OFF)
@@ -11,7 +11,8 @@ SET(CUDA_LIB "" CACHE PATH "Location of libraries")
 SET(CUDNN_LIB "" CACHE PATH "Location of libraries")
 SET(TENSORRT_DIR "" CACHE PATH "Compile demo with TensorRT")
 
-set(DEMO_NAME "ocr_system")
+set(DEMO_NAME "ppocr")
+
 
 macro(safe_set_static_flag)
     foreach(flag_var
@@ -205,6 +206,10 @@ endif()
 
 set(DEPS ${DEPS} ${OpenCV_LIBS})
 
+include(ExternalProject)
+include(external-cmake/auto-log.cmake)
+include_directories(${CMAKE_CURRENT_BINARY_DIR}/autolog/src/extern_Autolog/auto_log)
+
 AUX_SOURCE_DIRECTORY(./src SRCS)
 add_executable(${DEMO_NAME} ${SRCS})
 

deploy/cpp_infer/external-cmake/auto-log.cmake

+find_package(Git REQUIRED)
+message("${CMAKE_BUILD_TYPE}")
+
+set(AUTOLOG_REPOSITORY     https://github.com/LDOUBLEV/AutoLog.git)
+SET(AUTOLOG_INSTALL_DIR   ${CMAKE_CURRENT_BINARY_DIR}/install/Autolog)
+
+ExternalProject_Add(
+    extern_Autolog
+    PREFIX autolog
+    GIT_REPOSITORY ${AUTOLOG_REPOSITORY}
+    GIT_TAG main
+    DOWNLOAD_NO_EXTRACT True
+    INSTALL_COMMAND cmake -E echo "Skipping install step."
+)

deploy/cpp_infer/include/clipper.h

@@ -31,6 +31,8 @@
 *                                                                              *
 *******************************************************************************/
 
+#pragma once
+
 #ifndef clipper_hpp
 #define clipper_hpp
 

deploy/cpp_infer/include/config.h

-// 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.
-
-#pragma once
-
-#include <iomanip>
-#include <iostream>
-#include <map>
-#include <ostream>
-#include <string>
-#include <vector>
-
-#include "include/utility.h"
-
-namespace PaddleOCR {
-
-class OCRConfig {
-public:
-  explicit OCRConfig(const std::string &config_file) {
-    config_map_ = LoadConfig(config_file);
-
-    this->use_gpu = bool(stoi(config_map_["use_gpu"]));
-
-    this->gpu_id = stoi(config_map_["gpu_id"]);
-
-    this->gpu_mem = stoi(config_map_["gpu_mem"]);
-
-    this->cpu_math_library_num_threads =
-        stoi(config_map_["cpu_math_library_num_threads"]);
-
-    this->use_mkldnn = bool(stoi(config_map_["use_mkldnn"]));
-
-    this->max_side_len = stoi(config_map_["max_side_len"]);
-
-    this->det_db_thresh = stod(config_map_["det_db_thresh"]);
-
-    this->det_db_box_thresh = stod(config_map_["det_db_box_thresh"]);
-
-    this->det_db_unclip_ratio = stod(config_map_["det_db_unclip_ratio"]);
-
-    this->use_polygon_score = bool(stoi(config_map_["use_polygon_score"]));
-
-    this->det_model_dir.assign(config_map_["det_model_dir"]);
-
-    this->rec_model_dir.assign(config_map_["rec_model_dir"]);
-
-    this->char_list_file.assign(config_map_["char_list_file"]);
-
-    this->use_angle_cls = bool(stoi(config_map_["use_angle_cls"]));
-
-    this->cls_model_dir.assign(config_map_["cls_model_dir"]);
-
-    this->cls_thresh = stod(config_map_["cls_thresh"]);
-
-    this->visualize = bool(stoi(config_map_["visualize"]));
-
-    this->use_tensorrt = bool(stoi(config_map_["use_tensorrt"]));
-
-    this->use_fp16 = bool(stod(config_map_["use_fp16"]));
-  }
-
-  bool use_gpu = false;
-
-  int gpu_id = 0;
-
-  int gpu_mem = 4000;
-
-  int cpu_math_library_num_threads = 1;
-
-  bool use_mkldnn = false;
-
-  int max_side_len = 960;
-
-  double det_db_thresh = 0.3;
-
-  double det_db_box_thresh = 0.5;
-
-  double det_db_unclip_ratio = 2.0;
-
-  bool use_polygon_score = false;
-
-  std::string det_model_dir;
-
-  std::string rec_model_dir;
-
-  bool use_angle_cls;
-
-  std::string char_list_file;
-
-  std::string cls_model_dir;
-
-  double cls_thresh;
-
-  bool visualize = true;
-
-  bool use_tensorrt = false;
-
-  bool use_fp16 = false;
-
-  void PrintConfigInfo();
-
-private:
-  // Load configuration
-  std::map<std::string, std::string> LoadConfig(const std::string &config_file);
-
-  std::vector<std::string> split(const std::string &str,
-                                 const std::string &delim);
-
-  std::map<std::string, std::string> config_map_;
-};
-
-} // namespace PaddleOCR

deploy/cpp_infer/include/ocr_cls.h

@@ -12,6 +12,8 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
+#pragma once
+
 #include "opencv2/core.hpp"
 #include "opencv2/imgcodecs.hpp"
 #include "opencv2/imgproc.hpp"
@@ -40,7 +42,7 @@ public:
                       const int &gpu_id, const int &gpu_mem,
                       const int &cpu_math_library_num_threads,
                       const bool &use_mkldnn, const double &cls_thresh,
-                      const bool &use_tensorrt, const bool &use_fp16) {
+                      const bool &use_tensorrt, const std::string &precision) {
     this->use_gpu_ = use_gpu;
     this->gpu_id_ = gpu_id;
     this->gpu_mem_ = gpu_mem;
@@ -49,7 +51,7 @@ public:
 
     this->cls_thresh = cls_thresh;
     this->use_tensorrt_ = use_tensorrt;
-    this->use_fp16_ = use_fp16;
+    this->precision_ = precision;
 
     LoadModel(model_dir);
   }
@@ -73,7 +75,7 @@ private:
   std::vector<float> scale_ = {1 / 0.5f, 1 / 0.5f, 1 / 0.5f};
   bool is_scale_ = true;
   bool use_tensorrt_ = false;
-  bool use_fp16_ = false;
+  std::string precision_ = "fp32";
   // pre-process
   ClsResizeImg resize_op_;
   Normalize normalize_op_;

deploy/cpp_infer/include/ocr_det.h

@@ -46,7 +46,7 @@ public:
                       const double &det_db_box_thresh,
                       const double &det_db_unclip_ratio,
                       const bool &use_polygon_score, const bool &visualize,
-                      const bool &use_tensorrt, const bool &use_fp16) {
+                      const bool &use_tensorrt, const std::string &precision) {
     this->use_gpu_ = use_gpu;
     this->gpu_id_ = gpu_id;
     this->gpu_mem_ = gpu_mem;
@@ -62,7 +62,7 @@ public:
 
     this->visualize_ = visualize;
     this->use_tensorrt_ = use_tensorrt;
-    this->use_fp16_ = use_fp16;
+    this->precision_ = precision;
 
     LoadModel(model_dir);
   }
@@ -71,7 +71,7 @@ public:
   void LoadModel(const std::string &model_dir);
 
   // Run predictor
-  void Run(cv::Mat &img, std::vector<std::vector<std::vector<int>>> &boxes);
+  void Run(cv::Mat &img, std::vector<std::vector<std::vector<int>>> &boxes, std::vector<double> *times);
 
 private:
   std::shared_ptr<Predictor> predictor_;
@@ -91,7 +91,7 @@ private:
 
   bool visualize_ = true;
   bool use_tensorrt_ = false;
-  bool use_fp16_ = false;
+  std::string precision_ = "fp32";
 
   std::vector<float> mean_ = {0.485f, 0.456f, 0.406f};
   std::vector<float> scale_ = {1 / 0.229f, 1 / 0.224f, 1 / 0.225f};

deploy/cpp_infer/include/ocr_rec.h

@@ -12,6 +12,8 @@
 // See the License for the specific language governing permissions and
 // limitations under the License.
 
+#pragma once
+
 #include "opencv2/core.hpp"
 #include "opencv2/imgcodecs.hpp"
 #include "opencv2/imgproc.hpp"
@@ -42,14 +44,14 @@ public:
                           const int &gpu_id, const int &gpu_mem,
                           const int &cpu_math_library_num_threads,
                           const bool &use_mkldnn, const string &label_path,
-                          const bool &use_tensorrt, const bool &use_fp16) {
+                          const bool &use_tensorrt, const std::string &precision) {
     this->use_gpu_ = use_gpu;
     this->gpu_id_ = gpu_id;
     this->gpu_mem_ = gpu_mem;
     this->cpu_math_library_num_threads_ = cpu_math_library_num_threads;
     this->use_mkldnn_ = use_mkldnn;
     this->use_tensorrt_ = use_tensorrt;
-    this->use_fp16_ = use_fp16;
+    this->precision_ = precision;
 
     this->label_list_ = Utility::ReadDict(label_path);
     this->label_list_.insert(this->label_list_.begin(),
@@ -62,8 +64,7 @@ public:
   // Load Paddle inference model
   void LoadModel(const std::string &model_dir);
 
-  void Run(std::vector<std::vector<std::vector<int>>> boxes, cv::Mat &img,
-           Classifier *cls);
+  void Run(cv::Mat &img, std::vector<double> *times);
 
 private:
   std::shared_ptr<Predictor> predictor_;
@@ -80,7 +81,7 @@ private:
   std::vector<float> scale_ = {1 / 0.5f, 1 / 0.5f, 1 / 0.5f};
   bool is_scale_ = true;
   bool use_tensorrt_ = false;
-  bool use_fp16_ = false;
+  std::string precision_ = "fp32";
   // pre-process
   CrnnResizeImg resize_op_;
   Normalize normalize_op_;
@@ -89,9 +90,6 @@ private:
   // post-process
   PostProcessor post_processor_;
 
-  cv::Mat GetRotateCropImage(const cv::Mat &srcimage,
-                             std::vector<std::vector<int>> box);
-
 }; // class CrnnRecognizer
 
 } // namespace PaddleOCR

deploy/cpp_infer/include/utility.h

@@ -47,6 +47,9 @@ public:
 
   static void GetAllFiles(const char *dir_name,
                           std::vector<std::string> &all_inputs);
+    
+  static cv::Mat GetRotateCropImage(const cv::Mat &srcimage,
+                          std::vector<std::vector<int>> box);
 };
 
 } // namespace PaddleOCR
\ No newline at end of file