Upload preprocessing.py

preprocessing.py

@@ -0,0 +1,59 @@
+def load_dataset_images(directory):
+    import pickle
+    import os
+    import cv2
+    import numpy as np
+
+    # Set the paths to the dataset folders
+    dataset_dir = "general"
+    train_dir = os.path.join(dataset_dir, "PANTOLON")
+
+    # Set the image size to be used as the input data for the network
+    image_size = (224, 224)
+
+    preprocessed_images_file = os.path.join(directory, 'preprocessed_images.pkl')
+    labels_file = os.path.join(directory, 'labels.pkl')
+    filenames_file = os.path.join(directory, 'filenames.pkl')
+
+    # Check if preprocessed images, labels and filenames already exist
+    if os.path.exists(preprocessed_images_file) and os.path.exists(labels_file) and os.path.exists(filenames_file):
+        # Load preprocessed images, labels and filenames from disk
+        with open(preprocessed_images_file, 'rb') as f:
+            images = pickle.load(f)
+        with open(labels_file, 'rb') as f:
+            labels = pickle.load(f)
+        with open(filenames_file, 'rb') as f:
+            filenames = pickle.load(f)
+    else:
+        # Preprocess images, labels and filenames and save them to disk
+        images = []
+        labels = []
+        filenames = []
+
+        for filename in os.listdir(directory):
+            if filename.endswith(".txt"):
+                with open(os.path.join(directory, filename), "r", encoding="utf-8") as file:
+                    product = file.readline().strip().replace("Product: ", "")
+                    labels.append(product)
+
+            if filename.endswith(".jpg") or filename.endswith(".JPG") or filename.endswith(
+                    ".jpeg") or filename.endswith(".png"):
+                img_path = os.path.join(directory, filename)
+                img = cv2.imread(img_path)
+                img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
+                img = cv2.resize(img, image_size)
+                img = img.astype("float32") / 255.0  # Normalize the pixel values
+                images.append(img)
+                filenames.append(filename)
+
+        images = np.array(images)
+
+        # Save preprocessed images, labels and filenames to disk
+        with open(preprocessed_images_file, 'wb') as f:
+            pickle.dump(images, f)
+        with open(labels_file, 'wb') as f:
+            pickle.dump(labels, f)
+        with open(filenames_file, 'wb') as f:
+            pickle.dump(filenames, f)
+
+    return images, labels, filenames