Spaces:
Sleeping
Sleeping
File size: 8,272 Bytes
d2b859c | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 | """
Module for preprocessing and data augmentation of the BCCD dataset.
"""
import os
import shutil
import cv2
import numpy as np
import xml.etree.ElementTree as ET
from tqdm import tqdm
import albumentations as A
from PIL import Image
def preprocess_dataset(dataset_path):
"""
Preprocesses the BCCD dataset images.
Args:
dataset_path: Path to the BCCD dataset
Returns:
Path to the preprocessed dataset
"""
print("Preprocessing dataset...")
# Create output directory
output_dir = "preprocessed_dataset"
os.makedirs(output_dir, exist_ok=True)
os.makedirs(os.path.join(output_dir, "images"), exist_ok=True)
os.makedirs(os.path.join(output_dir, "annotations"), exist_ok=True)
# Get paths
image_path = os.path.join(dataset_path, "BCCD", "JPEGImages")
annot_path = os.path.join(dataset_path, "BCCD", "Annotations")
# Get all image files
image_files = [f for f in os.listdir(image_path) if f.endswith(('.jpg', '.jpeg', '.png'))]
for file in tqdm(image_files, desc="Preprocessing images"):
# Read image
img = cv2.imread(os.path.join(image_path, file))
if img is None:
continue
# Apply preprocessing
# 1. Convert to RGB
img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
# 2. Normalize
img = img / 255.0
# 3. Resize to a standard size if needed
img = cv2.resize(img, (640, 640))
# 4. Convert back to 0-255 range and BGR for saving
img = (img * 255).astype(np.uint8)
img = cv2.cvtColor(img, cv2.COLOR_RGB2BGR)
# Save the preprocessed image
cv2.imwrite(os.path.join(output_dir, "images", file), img)
# Copy the annotation file
base_name = os.path.splitext(file)[0]
xml_file = os.path.join(annot_path, base_name + ".xml")
if os.path.exists(xml_file):
shutil.copy(xml_file, os.path.join(output_dir, "annotations", base_name + ".xml"))
print(f"Preprocessing completed. Saved to {output_dir}")
return output_dir
def augment_dataset(dataset_path):
"""
Applies data augmentation to the preprocessed dataset.
Args:
dataset_path: Path to the preprocessed dataset
Returns:
Path to the augmented dataset
"""
print("Augmenting dataset...")
# Create output directory
output_dir = "augmented_dataset"
os.makedirs(output_dir, exist_ok=True)
os.makedirs(os.path.join(output_dir, "images"), exist_ok=True)
os.makedirs(os.path.join(output_dir, "annotations"), exist_ok=True)
# Copy original data first
image_path = os.path.join(dataset_path, "images")
annot_path = os.path.join(dataset_path, "annotations")
for file in os.listdir(image_path):
shutil.copy(os.path.join(image_path, file),
os.path.join(output_dir, "images", file))
for file in os.listdir(annot_path):
shutil.copy(os.path.join(annot_path, file),
os.path.join(output_dir, "annotations", file))
# Define augmentation pipeline
augmentations = [
A.Compose([
A.HorizontalFlip(p=1.0),
A.BBoxParams(format='pascal_voc', label_fields=['class_labels'])
]),
A.Compose([
A.RandomBrightnessContrast(p=1.0),
A.BBoxParams(format='pascal_voc', label_fields=['class_labels'])
]),
A.Compose([
A.Rotate(limit=20, p=1.0),
A.BBoxParams(format='pascal_voc', label_fields=['class_labels'])
]),
A.Compose([
A.RandomSizedBBoxSafeCrop(width=640, height=640, p=1.0),
A.BBoxParams(format='pascal_voc', label_fields=['class_labels'])
])
]
# Get all image files
image_files = [f for f in os.listdir(image_path) if f.endswith(('.jpg', '.jpeg', '.png'))]
for file in tqdm(image_files, desc="Augmenting images"):
# Read image
img_path = os.path.join(image_path, file)
img = cv2.imread(img_path)
if img is None:
continue
# Convert to RGB
img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
# Read annotation
base_name = os.path.splitext(file)[0]
xml_path = os.path.join(annot_path, base_name + ".xml")
if not os.path.exists(xml_path):
continue
# Parse XML to get bounding boxes
tree = ET.parse(xml_path)
root = tree.getroot()
bboxes = []
class_labels = []
for obj in root.findall('object'):
cls = obj.find('name').text
bbox = obj.find('bndbox')
xmin = int(float(bbox.find('xmin').text))
ymin = int(float(bbox.find('ymin').text))
xmax = int(float(bbox.find('xmax').text))
ymax = int(float(bbox.find('ymax').text))
bboxes.append([xmin, ymin, xmax, ymax])
class_labels.append(cls)
# Apply each augmentation
for i, aug in enumerate(augmentations):
# Apply augmentation
try:
augmented = aug(image=img, bboxes=bboxes, class_labels=class_labels)
aug_img = augmented['image']
aug_bboxes = augmented['bboxes']
aug_labels = augmented['class_labels']
# Skip if no bounding boxes are left after augmentation
if len(aug_bboxes) == 0:
continue
# Create a new XML file for the augmented image
new_file_name = f"{base_name}_aug_{i}.jpg"
new_xml_name = f"{base_name}_aug_{i}.xml"
# Save the augmented image
aug_img_bgr = cv2.cvtColor(aug_img, cv2.COLOR_RGB2BGR)
cv2.imwrite(os.path.join(output_dir, "images", new_file_name), aug_img_bgr)
# Create a new XML annotation
create_xml_annotation(
os.path.join(output_dir, "annotations", new_xml_name),
new_file_name,
aug_img.shape[1], # width
aug_img.shape[0], # height
aug_bboxes,
aug_labels
)
except Exception as e:
print(f"Error augmenting {file} with {i}: {e}")
continue
print(f"Augmentation completed. Saved to {output_dir}")
return output_dir
def create_xml_annotation(path, filename, width, height, bboxes, labels):
"""
Creates an XML annotation file in Pascal VOC format.
Args:
path: Path to save the XML file
filename: Image filename
width: Image width
height: Image height
bboxes: List of bounding boxes [xmin, ymin, xmax, ymax]
labels: List of class labels
"""
root = ET.Element("annotation")
# Add image information
folder = ET.SubElement(root, "folder")
folder.text = "images"
file_node = ET.SubElement(root, "filename")
file_node.text = filename
size = ET.SubElement(root, "size")
width_node = ET.SubElement(size, "width")
width_node.text = str(width)
height_node = ET.SubElement(size, "height")
height_node.text = str(height)
depth = ET.SubElement(size, "depth")
depth.text = "3"
# Add object information
for (xmin, ymin, xmax, ymax), label in zip(bboxes, labels):
obj = ET.SubElement(root, "object")
name = ET.SubElement(obj, "name")
name.text = label
bndbox = ET.SubElement(obj, "bndbox")
xmin_node = ET.SubElement(bndbox, "xmin")
xmin_node.text = str(int(xmin))
ymin_node = ET.SubElement(bndbox, "ymin")
ymin_node.text = str(int(ymin))
xmax_node = ET.SubElement(bndbox, "xmax")
xmax_node.text = str(int(xmax))
ymax_node = ET.SubElement(bndbox, "ymax")
ymax_node.text = str(int(ymax))
# Write to file
tree = ET.ElementTree(root)
tree.write(path)
|