散称训练数据前置处理

data_preprocessing/create_extra.py

@@ -0,0 +1,92 @@
+import os
+import shutil
+from pathlib import Path
+
+def count_files(directory):
+    """统计目录中的文件数量"""
+    try:
+        return len([f for f in os.listdir(directory)
+                   if os.path.isfile(os.path.join(directory, f))])
+    except Exception as e:
+        print(f"无法统计目录 {directory}: {e}")
+        return 0
+
+def clear_empty_dirs(path):
+    """
+    删除空目录
+    :param path: 目录路径
+    """
+    for root, dirs, files in os.walk(path, topdown=False):
+        for dir_name in dirs:
+            dir_path = os.path.join(root, dir_name)
+            try:
+                if not os.listdir(dir_path):
+                    os.rmdir(dir_path)
+                    print(f"Deleted empty directory: {dir_path}")
+            except Exception as e:
+                print(f"Error: {e.strerror}")
+
+def megre_subdirs(pth):
+    for roots, dir_names, files in os.walk(pth):
+        print(f"image  {dir_names}")
+        for image in dir_names:
+            inner_dir_path = os.path.join(pth, image)
+            for inner_roots, inner_dirs, inner_files in os.walk(inner_dir_path):
+                for inner_dir in inner_dirs:
+                    src_dir = os.path.join(inner_roots, inner_dir)
+                    dest_dir = os.path.join(pth, inner_dir)
+                    # shutil.copytree(src_dir, dest_dir)
+                    shutil.move(src_dir, dest_dir)
+                    print(f"Copied {inner_dir} to {pth}")
+    clear_empty_dirs(pth)
+
+
+def split_subdirs(source_dir, target_dir, max_files=10):
+    """
+    复制文件数≤max_files的子目录到目标目录
+    :param source_dir: 源目录路径
+    :param target_dir: 目标目录路径
+    :param max_files: 最大文件数阈值
+    """
+    megre_subdirs(source_dir)  # 合并子目录,删除上级目录
+    # 创建目标目录
+    Path(target_dir).mkdir(parents=True, exist_ok=True)
+
+    print(f"开始处理目录: {source_dir}")
+    print(f"目标目录: {target_dir}")
+    print(f"筛选条件: 文件数 ≤ {max_files}\n")
+
+    # 遍历源目录
+    for subdir in os.listdir(source_dir):
+        subdir_path = os.path.join(source_dir, subdir)
+
+        if not os.path.isdir(subdir_path):
+            continue
+
+        try:
+            file_count = count_files(subdir_path)
+
+            if file_count <= max_files:
+                print(f"复制 {subdir} (包含 {file_count} 个文件)")
+                dest_path = os.path.join(target_dir, subdir)
+
+                # 如果目标目录已存在则跳过
+                if os.path.exists(dest_path):
+                    print(f"目录已存在，跳过: {dest_path}")
+                    continue
+
+                # shutil.copytree(subdir_path, dest_path)
+                shutil.move(subdir_path, dest_path)
+
+        except Exception as e:
+            print(f"处理目录 {subdir} 时出错: {e}")
+
+    print("\n处理完成")
+
+if __name__ == "__main__":
+    # 配置路径
+    SOURCE_DIR = r"C:\Users\123\Desktop\test1\scatter_sub_class"
+    TARGET_DIR = "scatter_mini"
+
+    # 执行复制操作
+    split_subdirs(SOURCE_DIR, TARGET_DIR)