Refactor and enhance OCR functionality with PaddleOCR 3.0 support (#161)

* Refactor and enhance OCR functionality with PaddleOCR 3.0 support

- Updated PaddleOCR dependency to version 3.0 for improved performance and features.
- Refactored OCR-related functions to utilize the new PaddleOCR API, including changes in model initialization and prediction methods.
- Enhanced error handling and logging throughout the OCR process for better debugging and user feedback.
- Updated documentation strings for clarity and consistency.
- Removed deprecated code and unnecessary imports to streamline the codebase.
- Added new table recognition capabilities using the updated PaddleOCR features.

This commit aims to improve the overall functionality and maintainability of the OCR system.

* Refactor code for improved readability and consistency

- Cleaned up formatting in multiple files for better readability, including adjustments to line breaks and indentation.

* Update tablepyxl/__init__.py

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

* Update tablepyxl/LICENSE

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

* Update libs/shape.py

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

* Update libs/shape.py

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

* Update version information to 3.0.0 in __init__.py

---------

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Author: GreatV

PPOCRLabel.py

@@ -5,7 +5,7 @@
 import argparse
 
 
-# 删除划分的训练集、验证集、测试集文件夹，重新创建一个空的文件夹
+# Delete the divided train, val, and test folders and create a new empty folder
 def isCreateOrDeleteFolder(path, flag):
     flagPath = os.path.join(path, flag)
 
@@ -66,7 +66,7 @@ def splitTrainVal(
                 test_txt.write("{}\t{}".format(image_copy_path, image_label))
 
 
-# 删掉存在的文件
+# Remove the file if it exists
 def removeFile(path):
     if os.path.exists(path):
         os.remove(path)
@@ -128,9 +128,13 @@ def genDetRecTrainVal(args):
 
 
 if __name__ == "__main__":
-    # 功能描述：分别划分检测和识别的训练集、验证集、测试集
-    # 说明：可以根据自己的路径和需求调整参数，图像数据往往多人合作分批标注，每一批图像数据放在一个文件夹内用PPOCRLabel进行标注，
-    # 如此会有多个标注好的图像文件夹汇总并划分训练集、验证集、测试集的需求
+    """
+    Function description: Split detection and recognition datasets into training, validation, and test sets
+    Note: You can adjust parameters according to your own path and needs. Image data is often annotated
+    in batches by multiple people collaborating. Each batch of image data is placed in a folder and
+    annotated using PPOCRLabel. This creates a need to aggregate multiple annotated image folders
+    and split them into training, validation, and test sets.
+    """
     parser = argparse.ArgumentParser()
     parser.add_argument(
         "--trainValTestRatio",

gen_ocr_train_val_test.py

@@ -1,2 +1,2 @@
-__version_info__ = ("2", "1", "12")
+__version_info__ = ("3", "0", "0")
 __version__ = ".".join(__version_info__)

libs/__init__.py

@@ -1,59 +1,71 @@
-try:
-    from PyQt5.QtGui import *
-    from PyQt5.QtCore import *
-    from PyQt5.QtWidgets import *
-except ImportError:
-    from PyQt4.QtGui import *
-    from PyQt4.QtCore import *
-
-import time
 import datetime
 import json
+import logging
+import time
+
 import cv2
 import numpy as np
+from PyQt5.QtCore import QThread, pyqtSignal, Qt
+from PyQt5.QtWidgets import (
+    QDialog,
+    QDialogButtonBox as BB,
+    QProgressBar,
+    QVBoxLayout,
+    QListWidget,
+)
 
 from libs.utils import newIcon
 
-BB = QDialogButtonBox
+logger = logging.getLogger("PPOCRLabel")
 
 
 class Worker(QThread):
     progressBarValue = pyqtSignal(int)
     listValue = pyqtSignal(str)
-    endsignal = pyqtSignal(int, str)
+    end_signal = pyqtSignal(int, str)
     handle = 0
 
-    def __init__(self, ocr, mImgList, mainThread, model):
+    def __init__(self, ocr, img_list, main_thread, model):
         super(Worker, self).__init__()
+        self.result_dic = None
         self.ocr = ocr
-        self.mImgList = mImgList
-        self.mainThread = mainThread
+        self.img_list = img_list
+        self.mainThread = main_thread
         self.model = model
         self.setStackSize(1024 * 1024)
 
     def run(self):
         try:
             findex = 0
-            for Imgpath in self.mImgList:
+            for img_path in self.img_list:
                 if self.handle == 0:
-                    self.listValue.emit(Imgpath)
+                    self.listValue.emit(img_path)
                     if self.model == "paddle":
                         h, w, _ = cv2.imdecode(
-                            np.fromfile(Imgpath, dtype=np.uint8), 1
+                            np.fromfile(img_path, dtype=np.uint8), 1
                         ).shape
                         if h > 32 and w > 32:
-                            self.result_dic = self.ocr.ocr(Imgpath, cls=True, det=True)[
-                                0
-                            ]
+                            result = self.ocr.predict(img_path)[0]
+                            self.result_dic = []
+                            for poly, text, score in zip(
+                                result["rec_polys"],
+                                result["rec_texts"],
+                                result["rec_scores"],
+                            ):
+                                # Convert numpy array to list for JSON serialization
+                                poly_list = (
+                                    poly.tolist() if hasattr(poly, "tolist") else poly
+                                )
+                                self.result_dic.append([poly_list, (text, score)])
                         else:
-                            print(
-                                "The size of", Imgpath, "is too small to be recognised"
+                            logger.warning(
+                                "The size of %s is too small to be recognised", img_path
                             )
                             self.result_dic = None
 
                     # 结果保存
                     if self.result_dic is None or len(self.result_dic) == 0:
-                        print("Can not recognise file", Imgpath)
+                        logger.warning("Can not recognise file %s", img_path)
                         pass
                     else:
                         strs = ""
@@ -73,32 +85,37 @@ def run(self):
                         # Sending large amounts of data repeatedly through pyqtSignal may affect the program efficiency
                         self.listValue.emit(strs)
                         self.mainThread.result_dic = self.result_dic
-                        self.mainThread.filePath = Imgpath
+                        self.mainThread.filePath = img_path
                         # 保存
                         self.mainThread.saveFile(mode="Auto")
                     findex += 1
                     self.progressBarValue.emit(findex)
                 else:
                     break
-            self.endsignal.emit(0, "readAll")
+            self.end_signal.emit(0, "readAll")
             self.exec()
         except Exception as e:
-            print(e)
+            logger.error("Error in worker thread: %s", e)
             raise
 
 
 class AutoDialog(QDialog):
     def __init__(
-        self, text="Enter object label", parent=None, ocr=None, mImgList=None, lenbar=0
+        self,
+        text="Enter object label",
+        parent=None,
+        ocr=None,
+        image_list=None,
+        len_bar=0,
     ):
         super(AutoDialog, self).__init__(parent)
         self.setFixedWidth(1000)
         self.parent = parent
         self.ocr = ocr
-        self.mImgList = mImgList
-        self.lender = lenbar
-        self.pb = QProgressBar()
-        self.pb.setRange(0, self.lender)
+        self.img_list = image_list
+        self.len_bar = len_bar
+        self.pb = QProgressBar(parent)
+        self.pb.setRange(0, self.len_bar)
         self.pb.setValue(0)
 
         layout = QVBoxLayout()
@@ -121,25 +138,25 @@ def __init__(
 
         # self.setWindowFlags(Qt.WindowCloseButtonHint)
 
-        self.thread_1 = Worker(self.ocr, self.mImgList, self.parent, "paddle")
+        self.thread_1 = Worker(self.ocr, self.img_list, self.parent, "paddle")
         self.thread_1.progressBarValue.connect(self.handleProgressBarSingal)
         self.thread_1.listValue.connect(self.handleListWidgetSingal)
-        self.thread_1.endsignal.connect(self.handleEndsignalSignal)
+        self.thread_1.end_signal.connect(self.handleEndsignalSignal)
         self.time_start = time.time()  # save start time
 
     def handleProgressBarSingal(self, i):
         self.pb.setValue(i)
 
         # calculate time left of auto labeling
-        avg_time = (
-            time.time() - self.time_start
-        ) / i  # Use average time to prevent time fluctuations
-        time_left = str(datetime.timedelta(seconds=avg_time * (self.lender - i))).split(
-            "."
-        )[
+        # Use average time to prevent time fluctuations
+        avg_time = (time.time() - self.time_start) / i
+        time_left = str(
+            datetime.timedelta(seconds=avg_time * (self.len_bar - i))
+        ).split(".")[
             0
         ]  # Remove microseconds
-        self.setWindowTitle("PPOCRLabel  --  " + f"Time Left: {time_left}")  # show
+        # show
+        self.setWindowTitle("PPOCRLabel  --  " + f"Time Left: {time_left}")
 
     def handleListWidgetSingal(self, i):
         self.listWidget.addItem(i)
@@ -152,14 +169,9 @@ def handleEndsignalSignal(self, i, str):
             self.buttonBox.button(BB.Cancel).setEnabled(False)
 
     def reject(self):
-        print("reject")
+        logger.debug("Auto recognition dialog rejected")
         self.thread_1.handle = -1
         self.thread_1.quit()
-        # del self.thread_1
-        # if self.thread_1.isRunning():
-        #     self.thread_1.terminate()
-        # self.thread_1.quit()
-        # super(AutoDialog,self).reject()
         while not self.thread_1.isFinished():
             pass
         self.accept()
@@ -170,22 +182,13 @@ def validate(self):
     def postProcess(self):
         try:
             self.edit.setText(self.edit.text().trimmed())
-            # print(self.edit.text())
         except AttributeError:
-            # PyQt5: AttributeError: 'str' object has no attribute 'trimmed'
             self.edit.setText(self.edit.text())
-            print(self.edit.text())
+            logger.debug("Auto dialog text: %s", self.edit.text())
 
     def popUp(self):
         self.thread_1.start()
         return 1 if self.exec_() else None
 
-    def closeEvent(self, event):
-        print("???")
-        # if self.thread_1.isRunning():
-        #     self.thread_1.quit()
-        #
-        #     # self._thread.terminate()
-        # # del self.thread_1
-        # super(AutoDialog, self).closeEvent(event)
+    def closeEvent(self, event, **kwargs):
         self.reject()

libs/autoDialog.py

@@ -12,13 +12,16 @@
 # THE SOFTWARE.
 
 import copy
+import logging
 
 from PyQt5.QtCore import Qt, pyqtSignal, QPointF, QPoint
 from PyQt5.QtGui import QPainter, QBrush, QColor, QPixmap
 from PyQt5.QtWidgets import QWidget, QMenu, QApplication
 from libs.shape import Shape
 from libs.utils import distance
 
+logger = logging.getLogger("PPOCRLabel")
+
 CURSOR_DEFAULT = Qt.ArrowCursor
 CURSOR_POINT = Qt.PointingHandCursor
 CURSOR_DRAW = Qt.CrossCursor
@@ -350,14 +353,16 @@ def handleDrawing(self, pos):
             if self.fourpoint:
                 targetPos = self.line[self.pointnum]
                 self.current.addPoint(targetPos)
-                print("current points in handleDrawing is ", self.line[self.pointnum])
+                logger.debug(
+                    "current points in handleDrawing is %s", self.line[self.pointnum]
+                )
                 self.update()
                 if self.pointnum == 3:
                     self.finalise()
 
             else:
                 initPos = self.current[0]
-                print("initPos", self.current[0])
+                logger.debug("initPos %s", self.current[0])
                 minX = initPos.x()
                 minY = initPos.y()
                 targetPos = self.line[1]
@@ -369,7 +374,7 @@ def handleDrawing(self, pos):
                 self.finalise()
 
         elif not self.outOfPixmap(pos):
-            print("release")
+            logger.debug("release")
             self.current = Shape()
             self.current.addPoint(pos)
             self.line.points = [pos, pos]
@@ -610,7 +615,7 @@ def paintEvent(self, event):
             and self.current is not None
             and len(self.current.points) >= 2
         ):
-            print("paint event")
+            logger.debug("paint event")
             drawing_shape = self.current.copy()
             drawing_shape.addPoint(self.line[1])
             drawing_shape.fill = True
@@ -737,7 +742,7 @@ def keyPressEvent(self, ev):
         self.shapesBackups.pop()
         self.shapesBackups.append(shapesBackup)
         if key == Qt.Key_Escape and self.current:
-            print("ESC press")
+            logger.debug("ESC press")
             self.current = None
             self.drawingPolygon.emit(False)
             self.update()

libs/canvas.py

@@ -10,15 +10,8 @@
 # SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF
 # CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
 # THE SOFTWARE.
-try:
-    from PyQt5.QtGui import *
-    from PyQt5.QtCore import *
-    from PyQt5.QtWidgets import QColorDialog, QDialogButtonBox
-except ImportError:
-    from PyQt4.QtGui import *
-    from PyQt4.QtCore import *
 
-BB = QDialogButtonBox
+from PyQt5.QtWidgets import QColorDialog, QDialogButtonBox as BB
 
 
 class ColorDialog(QColorDialog):

libs/colorDialog.py

@@ -13,11 +13,13 @@
 #!/usr/bin/env python
 # -*- coding: utf8 -*-
 import json
+import logging
 from pathlib import Path
-
 from libs.constants import DEFAULT_ENCODING
 import os
 
+logger = logging.getLogger("PPOCRLabel")
+
 JSON_EXT = ".json"
 ENCODE_METHOD = DEFAULT_ENCODING
 
@@ -114,7 +116,7 @@ def __init__(self, jsonpath, filepath):
         try:
             self.parse_json()
         except ValueError:
-            print("JSON decoding failed")
+            logger.error("JSON decoding failed")
 
     def parse_json(self):
         with open(self.jsonpath, "r") as file:

libs/create_ml_io.py

@@ -13,7 +13,7 @@ def item_clicked(self, modelindex: QModelIndex):
         try:
             if self.edited_item is not None:
                 self.closePersistentEditor(self.edited_item)
-        except:
+        except Exception:
             self.edited_item = self.currentItem()
 
         self.edited_item = self.item(modelindex.row())

libs/editinlist.py

@@ -12,25 +12,7 @@
 # THE SOFTWARE.
 #!/usr/bin/env python
 # -*- coding: utf-8 -*-
-import sys
-
-try:
-    from PyQt5.QtGui import *
-    from PyQt5.QtCore import *
-    from PyQt5.QtWidgets import *
-except ImportError:
-    # needed for py3+qt4
-    # Ref:
-    # http://pyqt.sourceforge.net/Docs/PyQt4/incompatible_apis.html
-    # http://stackoverflow.com/questions/21217399/pyqt4-qtcore-qvariant-object-instead-of-a-string
-    if sys.version_info.major >= 3:
-        import sip
-
-        sip.setapi("QVariant", 2)
-    from PyQt4.QtGui import *
-    from PyQt4.QtCore import *
-
-# PyQt5: TypeError: unhashable type: 'QListWidgetItem'
+from PyQt5.QtWidgets import QListWidgetItem
 
 
 class HashableQListWidgetItem(QListWidgetItem):