rename files

src/utils/utils.py

@@ -1,5 +1,8 @@
 import os
 import laspy
+import shutil
+import random
+import numpy as np
 
 def average_points_in_laz_files(directory):
     total_points = 0
@@ -24,6 +27,160 @@ def average_points_in_laz_files(directory):
     else:
         return "No LAZ files found or readable."
 
-directory = "data/bb_extracted_data_labelled"
-average = average_points_in_laz_files(directory)
-print(f"Average number of points per point cloud: {average}")
+def modify_labels(input_folder, output_folder):
+    """
+    Modifies the 'label' scalar field in LAZ files within the specified input folder and 
+    saves the modified point clouds to a new folder. This function changes the 'label' value 
+    from 2 to 1, and from 1 to 0, then writes the modified point cloud to a new LAZ file in the output folder.
+
+    Parameters:
+    input_folder (str): Path to the folder containing the original LAZ files.
+    output_folder (str): Path to the folder where modified LAZ files will be saved.
+    """
+
+    if not os.path.exists(output_folder):
+        os.makedirs(output_folder)
+
+    for filename in os.listdir(input_folder):
+        if filename.endswith(".laz"):
+            input_file_path = os.path.join(input_folder, filename)
+            output_file_path = os.path.join(output_folder, filename)
+
+            with laspy.open(input_file_path) as file:
+                las = file.read()
+
+                # Check if 'label' field exists
+                if 'label' in las.point_format.dimension_names:
+                    # Modify labels
+                    labels = las.label
+                    labels[labels == 2] = 1
+                    labels[labels == 1] = 0
+                    las.label = labels
+
+                    # Write to new file
+                    las.write(output_file_path)
+                else:
+                    print(f"'label' field not found in file {filename}")
+
+def split_dataset_into_folders(base_folder, train_ratio=0.8, val_ratio=0.1):
+    """
+    Splits the files in a folder into train, validation, and test folders 
+    according to specified ratios.
+
+    Parameters:
+    base_folder (str): Path to the base folder containing the files to split.
+    train_ratio (float): Proportion of files to be used for training.
+    val_ratio (float): Proportion of files to be used for validation.
+
+    The test ratio is automatically calculated as (1 - train_ratio - val_ratio).
+    """
+
+    # Paths for train, validation, and test folders
+    train_folder = os.path.join(base_folder, 'train')
+    val_folder = os.path.join(base_folder, 'val')
+    test_folder = os.path.join(base_folder, 'test')
+
+    # Create directories if they don't exist
+    for folder in [train_folder, val_folder, test_folder]:
+        if not os.path.exists(folder):
+            os.makedirs(folder)
+
+    # List all files in the base folder
+    all_files = [f for f in os.listdir(base_folder) if os.path.isfile(os.path.join(base_folder, f))]
+    random.shuffle(all_files)  # Shuffle to randomize file selection
+
+    # Calculate the number of files for train, val, and test
+    num_files = len(all_files)
+    num_train = int(num_files * train_ratio)
+    num_val = int(num_files * val_ratio)
+    num_test = num_files - num_train - num_val
+
+    # Split files into train, val, and test
+    for i, file in enumerate(all_files):
+        src_file = os.path.join(base_folder, file)
+
+        if i < num_train:
+            dest_folder = train_folder
+        elif i < num_train + num_val:
+            dest_folder = val_folder
+        else:
+            dest_folder = test_folder
+
+        shutil.move(src_file, os.path.join(dest_folder, file))
+
+def is_point_cloud_match(file1, file2):
+    """
+    Compares two LAZ files based on their xyz coordinates and RGB values.
+    Returns True if they match, False otherwise.
+    """
+    with laspy.open(file1) as f1, laspy.open(file2) as f2:
+        pc1, pc2 = f1.read(), f2.read()
+
+        # Compare xyz coordinates
+        coords_match = np.array_equal(pc1.xyz, pc2.xyz)
+
+        # Check if RGB data exists and compare it
+        rgb_match = True  # Assume true if RGB data is not present
+        if all(hasattr(pc1, color) and hasattr(pc2, color) for color in ['red', 'green', 'blue']):
+            rgb1 = np.vstack((pc1.red, pc1.green, pc1.blue)).T
+            rgb2 = np.vstack((pc2.red, pc2.green, pc2.blue)).T
+            rgb_match = np.array_equal(rgb1, rgb2)
+
+        return coords_match and rgb_match
+
+def copy_distinct_laz_files(folder1, folder2, output_folder):
+    """
+    Copies LAZ files from folder2 to output_folder if they are distinct
+    (in terms of xyz and RGB) from all files in folder1.
+    """
+    if not os.path.exists(output_folder):
+        os.makedirs(output_folder)
+
+    for file2 in os.listdir(folder2):
+        if file2.endswith(".laz"):
+            file2_path = os.path.join(folder2, file2)
+            match_found = False
+
+            for file1 in os.listdir(folder1):
+                if file1.endswith(".laz"):
+                    file1_path = os.path.join(folder1, file1)
+                    if is_point_cloud_match(file1_path, file2_path):
+                        match_found = True
+                        break
+
+            if not match_found:
+                shutil.copy(file2_path, os.path.join(output_folder, file2))
+
+# Example usage
+# folder1 = 'data/bb_extracted_data_labelled/bb_with_rgb'
+# folder2 = 'data/bb_extracted_data_unlabelled'
+# output_folder = 'data/bb_extracted_data_labelled/bb_to_label'
+# copy_distinct_laz_files(folder1, folder2, output_folder)
+
+# # Example usage
+# base_folder_path = 'data/bb_extracted_data_labelled'
+# split_dataset_into_folders(base_folder_path)
+
+def rename_files_to_pattern(directory):
+    """
+    Renames all files in the specified directory to 'bb_<number>.laz',
+    where <number> starts at 1 and increments for each file.
+    """
+    # Get a list of all files in the directory
+    files = [f for f in os.listdir(directory) if os.path.isfile(os.path.join(directory, f))]
+
+    # Iterate through the files and rename them
+    for index, file_name in enumerate(files, start=1):
+        new_file_name = f'bb_{index}.laz'
+        original_file_path = os.path.join(directory, file_name)
+        new_file_path = os.path.join(directory, new_file_name)
+
+        try:
+            os.rename(original_file_path, new_file_path)
+            print(f"Renamed: {file_name} to {new_file_name}")
+        except Exception as e:
+            print(f"Error renaming {file_name}: {e}")
+
+# Usage
+directory = 'data/bb_extracted_data_labelled/full dataset'  # Replace with your directory path
+rename_files_to_pattern(directory)