Merge remote-tracking branch 'origin/dygraph' into dygraph

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("..")
@@ -267,6 +272,8 @@ class MainWindow(QMainWindow, WindowMixin):
         self.colorDialog = ColorDialog(parent=self)
         self.zoomWidgetValue = self.zoomWidget.value()
 
+        self.msgBox = QMessageBox()
+
         ########## thumbnail #########
         hlayout = QHBoxLayout()
         m = (0, 0, 0, 0)
@@ -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)
 
@@ -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"), self.rotateLeftImg,
+                               'Ctrl+Alt+L', "rotateLeft", getStr("rotateLeft"), enabled=False)
+
+        rotateRight = action(getStr("rotateRight"), self.rotateRightImg,
+                               '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 +795,48 @@ 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 rotateLeftImg(self, _value=False):
+
+        filename = self.mImgList[self.currIndex]
+
+        if os.path.exists(filename):
+            if self.itemsToShapesbox:
+                self.rotateImgWarn()
+            else:
+                self.rotateImg(filename=filename, k=1, _value=True)
+        else:
+            self.rotateImgWarn()
+            self.actions.rotateRight.setEnabled(False)
+
+    def rotateRightImg(self, _value=False):
+
+        filename = self.mImgList[self.currIndex]
+
+        if os.path.exists(filename):
+            if self.itemsToShapesbox:
+                self.rotateImgWarn()
+            else:
+                self.rotateImg(filename=filename, k=-1, _value=True)
+        else:
+            self.rotateImgWarn()
+            self.actions.rotateRight.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 +944,12 @@ 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))
@@ -1418,6 +1482,7 @@ class MainWindow(QMainWindow, WindowMixin):
 
     def loadRecent(self, filename):
         if self.mayContinue():
+            print(filename,"======")
             self.loadFile(filename)
 
     def scanAllImages(self, folderPath):
@@ -1453,6 +1518,22 @@ class MainWindow(QMainWindow, WindowMixin):
         self.lastOpenDir = targetDirPath
         self.importDirImages(targetDirPath)
 
+    def openDatasetDirDialog(self,):
+        if not self.mayContinue():
+            return
+
+        if self.lastOpenDir and os.path.exists(self.lastOpenDir):
+            os.startfile(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 +1581,10 @@ 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):

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
@@ -709,17 +715,21 @@ 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)
+        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)
@@ -744,6 +754,9 @@ class Canvas(QWidget):
                 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 +853,10 @@ class Canvas(QWidget):
     def restoreShape(self):
         if not self.isShapeRestorable:
             return
+        if self.selectCountShape:
+            if len(self.shapesBackups) > 2:
+                for i in range(1,self.selectCount):
+                    self.shapesBackups.pop()
         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
@@ -85,6 +86,8 @@ detectionBoxposition=检测框位置
 recognitionResult=识别结果
 creatPolygon=四点标注
 drawSquares=正方形标注
+rotateLeft=图片左旋转90度
+rotateRight=图片右旋转90度
 saveRec=保存识别结果
 tempLabel=待识别
 nullLabel=无法识别

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
@@ -84,6 +85,8 @@ 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
 tempLabel=TEMPORARY

requirements.txt

@@ -7,4 +7,4 @@ tqdm
 numpy
 visualdl
 python-Levenshtein
-opencv-contrib-python==4.2.0.32
\ No newline at end of file
+opencv-contrib-python==4.4.0.46
\ No newline at end of file

tools/infer/utility.py

@@ -51,7 +51,7 @@ def init_args():
     parser.add_argument("--det_db_box_thresh", type=float, default=0.6)
     parser.add_argument("--det_db_unclip_ratio", type=float, default=1.5)
     parser.add_argument("--max_batch_size", type=int, default=10)
-    parser.add_argument("--use_dilation", type=bool, default=False)
+    parser.add_argument("--use_dilation", type=str2bool, default=False)
     parser.add_argument("--det_db_score_mode", type=str, default="fast")
     # EAST parmas
     parser.add_argument("--det_east_score_thresh", type=float, default=0.8)
@@ -61,7 +61,7 @@ def init_args():
     # SAST parmas
     parser.add_argument("--det_sast_score_thresh", type=float, default=0.5)
     parser.add_argument("--det_sast_nms_thresh", type=float, default=0.2)
-    parser.add_argument("--det_sast_polygon", type=bool, default=False)
+    parser.add_argument("--det_sast_polygon", type=str2bool, default=False)
 
     # params for text recognizer
     parser.add_argument("--rec_algorithm", type=str, default='CRNN')
@@ -90,7 +90,7 @@ def init_args():
     parser.add_argument(
         "--e2e_char_dict_path", type=str, default="./ppocr/utils/ic15_dict.txt")
     parser.add_argument("--e2e_pgnet_valid_set", type=str, default='totaltext')
-    parser.add_argument("--e2e_pgnet_polygon", type=bool, default=True)
+    parser.add_argument("--e2e_pgnet_polygon", type=str2bool, default=True)
     parser.add_argument("--e2e_pgnet_mode", type=str, default='fast')
 
     # params for text classifier
@@ -111,7 +111,7 @@ def init_args():
     parser.add_argument("--total_process_num", type=int, default=1)
     parser.add_argument("--process_id", type=int, default=0)
 
-    parser.add_argument("--benchmark", type=bool, default=False)
+    parser.add_argument("--benchmark", type=str2bool, default=False)
     parser.add_argument("--save_log_path", type=str, default="./log_output/")
 
     parser.add_argument("--show_log", type=str2bool, default=True)
@@ -210,22 +210,22 @@ def create_predictor(args, mode, logger):
                 "nearest_interp_v2_0.tmp_0": [1, 256, 40, 40]
             }
             min_pact_shape = {
-                "nearest_interp_v2_26.tmp_0":[1,256,20,20],
-                "nearest_interp_v2_27.tmp_0":[1,64,20,20],
-                "nearest_interp_v2_28.tmp_0":[1,64,20,20],
-                "nearest_interp_v2_29.tmp_0":[1,64,20,20]
+                "nearest_interp_v2_26.tmp_0": [1, 256, 20, 20],
+                "nearest_interp_v2_27.tmp_0": [1, 64, 20, 20],
+                "nearest_interp_v2_28.tmp_0": [1, 64, 20, 20],
+                "nearest_interp_v2_29.tmp_0": [1, 64, 20, 20]
             }
             max_pact_shape = {
-                "nearest_interp_v2_26.tmp_0":[1,256,400,400],
-                "nearest_interp_v2_27.tmp_0":[1,64,400,400],
-                "nearest_interp_v2_28.tmp_0":[1,64,400,400],
-                "nearest_interp_v2_29.tmp_0":[1,64,400,400]
+                "nearest_interp_v2_26.tmp_0": [1, 256, 400, 400],
+                "nearest_interp_v2_27.tmp_0": [1, 64, 400, 400],
+                "nearest_interp_v2_28.tmp_0": [1, 64, 400, 400],
+                "nearest_interp_v2_29.tmp_0": [1, 64, 400, 400]
             }
             opt_pact_shape = {
-                "nearest_interp_v2_26.tmp_0":[1,256,160,160],
-                "nearest_interp_v2_27.tmp_0":[1,64,160,160],
-                "nearest_interp_v2_28.tmp_0":[1,64,160,160],
-                "nearest_interp_v2_29.tmp_0":[1,64,160,160]
+                "nearest_interp_v2_26.tmp_0": [1, 256, 160, 160],
+                "nearest_interp_v2_27.tmp_0": [1, 64, 160, 160],
+                "nearest_interp_v2_28.tmp_0": [1, 64, 160, 160],
+                "nearest_interp_v2_29.tmp_0": [1, 64, 160, 160]
             }
             min_input_shape.update(min_pact_shape)
             max_input_shape.update(max_pact_shape)