Tldr 470 added full angle info

dedoc/readers/pdf_reader/pdf_image_reader/table_recognizer/table_utils/img_processing.py

@@ -4,26 +4,27 @@
 from typing import Any, List, Tuple
 
 import cv2
+import imutils
 import numpy as np
 
 from dedoc.config import get_config
 from dedoc.readers.pdf_reader.data_classes.tables.table_tree import TableTree
 from dedoc.readers.pdf_reader.data_classes.tables.table_type import TableTypeAdditionalOptions
-from dedoc.utils.image_utils import rotate_image
 
 logger = get_config().get("logger", logging.getLogger())
 logger = logger if logger else logging.getLogger("TableRecognizer.detect_tables_by_contours")
 table_options = TableTypeAdditionalOptions()
 
 
 def rotate_with_threshold(img: np.ndarray, angle: float, threshold: float = None, *, config: dict) -> np.ndarray:
+    """rotates a table image and saving image.shape during rotation. It is important for word bounding box extraction"""
     if threshold is None:
         threshold = config["rotate_threshold"]
     rotated = img
     if abs(angle) > threshold:
         if config.get("debug_mode", False):
             logger.debug("rotated image")
-        rotated = rotate_image(img, angle)
+        rotated = imutils.rotate(img, angle)
     return rotated
 
 

dedoc/scripts/test_words_bbox_extraction.py

@@ -118,7 +118,7 @@ def draw_word_annotations(self, image: np.ndarray, word_annotations: List[BboxWi
             p2 = (int((bbox["x_top_left"] + bbox["width"]) * bbox["page_width"]), int((bbox["y_top_left"] + bbox["height"]) * bbox["page_height"]))
 
             if angle != 0.0:
-                p1 = self.rotate_coordinate(p1[0], p1[1], x_c, y_c, angle)  # TODO x_c, y_c нужен четкий
+                p1 = self.rotate_coordinate(p1[0], p1[1], x_c, y_c, angle)
                 p2 = self.rotate_coordinate(p2[0], p2[1], x_c, y_c, angle)
 
             cv2.rectangle(image, p1, p2, (0, 255, 0) if ann.text_type == "typewritten" else (255, 0, 0))
@@ -155,10 +155,12 @@ def test_tabby_document(self):
         cv2.imwrite(os.path.join(output_path, f"{os.path.split(file_name)[1]}.png"), image)
 
     def test_table_word_extraction(self):
-        output_path = os.path.join(self.output_path)
+        output_path = os.path.join(self.output_path, 'tables')
         os.makedirs(output_path, exist_ok=True)
         file_names = ["tables/example_with_table5.png", "tables/example_with_table3.png", "tables/example_with_table4.jpg",
-                      "tables/example_with_table6.png", "tables/example_with_table_horizontal_union.jpg"]
+                      "tables/example_with_table6.png", "tables/example_with_table_horizontal_union.jpg",
+                      "scanned/orient_1.png"]
+
         for file_name in file_names:
             result = self._send_request(file_name, data=dict())
             table0 = result["content"]["tables"][0]
@@ -170,5 +172,5 @@ def test_table_word_extraction(self):
             image = rotate_image(image, page_angle)
 
             image = self.draw_word_annotations(image, word_annotations, angle=table_angle)
-            cv2.imwrite(os.path.join(output_path, file_name), image)
+            cv2.imwrite(os.path.join(output_path, file_name.split('/')[-1]), image)
 

dedoc/utils/image_utils.py

@@ -39,7 +39,8 @@ def get_bbox_from_image(image: Image, bbox: BBox, resize: Tuple[int, int] = (300
 
 def rotate_image(image: np.ndarray, angle: float, color_bound: Tuple[int, int, int] = (255, 255, 255)) -> np.ndarray:
     """
-    Rotates an image (angle in degrees) and expands image to avoid cropping
+    Rotates an image (angle in degrees) and expands image to avoid cropping (do bounds of color_bound)
+    Changes width and height of image (image.shape != rotated_image.shape)
     """
     height, width = image.shape[:2]
     image_center = (width / 2, height / 2)