使用面积占比方式判断一行文本是不是在一个layoutbox里

demo/ocr_demo.py

@@ -92,5 +92,5 @@ if __name__ == '__main__':
     ocr_json_file_path = r"/home/cxu/workspace/Magic-PDF/ocr_demo/j.1540-627x.2006.00176.x.json"
     # ocr_pdf_path = r"/home/cxu/workspace/Magic-PDF/ocr_demo/ocr_1.pdf"
     # ocr_json_file_path = r"/home/cxu/workspace/Magic-PDF/ocr_demo/ocr_1.json"
-    #ocr_local_parse(ocr_pdf_path, ocr_json_file_path)
-    ocr_online_parse(book_name="美国加州中学教材/edu_00000060")
+    ocr_local_parse(ocr_pdf_path, ocr_json_file_path)
+    #ocr_online_parse(book_name="美国加州中学教材/edu_00000060")

magic_pdf/dict2md/ocr_mkcontent.py

@@ -71,25 +71,26 @@ def ocr_mk_mm_markdown(pdf_info_dict: dict):
 def ocr_mk_mm_markdown_with_para(pdf_info_dict: dict):
     markdown = []
     for _, page_info in pdf_info_dict.items():
-        paras = page_info.get("para_blocks")
-        if not paras:
+        paras_of_layout = page_info.get("para_blocks")
+        if not paras_of_layout:
             continue
-        for para in paras:
-            para_text = ''
-            for line in para:
-                for span in line['spans']:
-                    span_type = span.get('type')
-                    if span_type == ContentType.Text:
-                        content = split_long_words(span['content'])
-                        # content = span['content']
-                    elif span_type == ContentType.InlineEquation:
-                        content = f"${span['content']}$"
-                    elif span_type == ContentType.InterlineEquation:
-                        content = f"\n$$\n{span['content']}\n$$\n"
-                    elif span_type in [ContentType.Image, ContentType.Table]:
-                        content = f"\n![]({join_path(s3_image_save_path, span['image_path'])})\n"
-                    para_text += content + ' '
-            markdown.append(para_text.strip() + '  ')
+        for paras in paras_of_layout:
+            for para in paras:
+                para_text = ''
+                for line in para:
+                    for span in line['spans']:
+                        span_type = span.get('type')
+                        if span_type == ContentType.Text:
+                            content = split_long_words(span['content'])
+                            pass
+                        elif span_type == ContentType.InlineEquation:
+                            content = f" ${span['content']}$ "
+                        elif span_type == ContentType.InterlineEquation:
+                            content = f"\n$$\n{span['content']}\n$$\n"
+                        elif span_type in [ ContentType.Image, ContentType.Table ]:
+                            content = f"\n![]({join_path(s3_image_save_path, span['image_path'])})\n"
+                        para_text += content + ' '
+                markdown.append(para_text.strip() + '  ')
 
     return '\n\n'.join(markdown)
 

magic_pdf/libs/boxbase.py

@@ -18,6 +18,33 @@ def _is_in_or_part_overlap(box1, box2) -> bool:
                 y1_1 < y0_2 or  # box1在box2的上边
                 y0_1 > y1_2)    # box1在box2的下边
 
+def _is_in_or_part_overlap_with_area_ratio(box1, box2, area_ratio_threshold=0.6):
+    """
+    判断box1是否在box2里面，或者box1和box2有部分重叠，且重叠面积占box1的比例超过area_ratio_threshold
+    
+    """
+    if box1 is None or box2 is None:
+        return False
+    
+    x0_1, y0_1, x1_1, y1_1 = box1
+    x0_2, y0_2, x1_2, y1_2 = box2
+
+    if not _is_in_or_part_overlap(box1, box2):
+        return False
+    
+    # 计算重叠面积
+    x_left = max(x0_1, x0_2)
+    y_top = max(y0_1, y0_2)
+    x_right = min(x1_1, x1_2)
+    y_bottom = min(y1_1, y1_2)
+    overlap_area = (x_right - x_left) * (y_bottom - y_top)
+    
+    # 计算box1的面积
+    box1_area = (x1_1 - x0_1) * (y1_1 - y0_1)
+    
+    return overlap_area / box1_area > area_ratio_threshold
+    
+    
 def _is_in(box1, box2) -> bool:
     """
     box1是否完全在box2里面