Initial commit

.gitignore

+images
+similar_images
+unique_images

datasort.py

+import os
+import cv2
+import shutil
+
+# Define the threshold for darkness detection
+darkness_threshold = 18  # You can adjust this value based on your needs
+
+# Define the threshold for blur detection
+blur_threshold = 100  # You can adjust this value based on your needs
+
+# Get the directory where the script is located
+script_directory = os.path.dirname(__file__)
+
+# Create directories to store the blurry, non-blurry, and dark images
+blurry_output_directory = os.path.join(script_directory, 'blurry_images')
+non_blurry_output_directory = os.path.join(script_directory, 'non_blurry_images')
+dark_output_directory = os.path.join(script_directory, 'dark_images')
+
+# Function to empty a folder
+def empty_folder(folder):
+    for filename in os.listdir(folder):
+        file_path = os.path.join(folder, filename)
+        try:
+            if os.path.isfile(file_path):
+                os.unlink(file_path)
+        except Exception as e:
+            print(e)
+
+# Empty the blurry, non-blurry, and dark image folders if they are non-empty
+if os.path.exists(blurry_output_directory):
+    empty_folder(blurry_output_directory)
+if os.path.exists(non_blurry_output_directory):
+    empty_folder(non_blurry_output_directory)
+if os.path.exists(dark_output_directory):
+    empty_folder(dark_output_directory)
+
+# Use os.walk to traverse all subdirectories
+for folder, subfolders, files in os.walk(script_directory):
+    for file in files:
+        # Check if the file is an image (you can customize the extensions)
+        if file.lower().endswith(('.jpg', '.jpeg', '.png', '.gif', '.bmp')):
+            image_path = os.path.join(folder, file)
+            
+            # Load the image
+            image = cv2.imread(image_path)
+            
+            # Calculate the mean pixel value to detect darkness
+            brightness = cv2.mean(image)[0]
+            
+            # Calculate the variance of Laplacian to detect blur
+            variance = cv2.Laplacian(image, cv2.CV_64F).var()
+            
+            # Create the output directory based on image quality (dark, blurry, or non-dark and non-blurry)
+            if brightness < darkness_threshold:
+                # Copy the dark image to the dark images folder
+                print(f"Dark Image: {image_path} - Brightness: {brightness}")
+                if not os.path.exists(dark_output_directory):
+                    os.makedirs(dark_output_directory)
+                destination_path = os.path.join(dark_output_directory, os.path.basename(image_path))
+                # Check if the source and destination paths are different
+                if image_path != destination_path:
+                    shutil.copy2(image_path, destination_path)
+            elif variance < blur_threshold:
+                # Copy the blurry image to the blurry images folder
+                print(f"Blurry Image: {image_path} - Variance: {variance}")
+                if not os.path.exists(blurry_output_directory):
+                    os.makedirs(blurry_output_directory)
+                destination_path = os.path.join(blurry_output_directory, os.path.basename(image_path))
+                # Check if the source and destination paths are different
+                if image_path != destination_path:
+                    shutil.copy2(image_path, destination_path)
+            else:
+                # Copy the non-dark and non-blurry image to the non-blurry images folder
+                print(f"Non-Dark and Non-Blurry Image: {image_path}")
+                if not os.path.exists(non_blurry_output_directory):
+                    os.makedirs(non_blurry_output_directory)
+                destination_path = os.path.join(non_blurry_output_directory, os.path.basename(image_path))
+                # Check if the source and destination paths are different
+                if image_path != destination_path:
+                    shutil.copy2(image_path, destination_path)
+
+# You can adjust the 'darkness_threshold', 'blur_threshold', 'blurry_output_directory', 'dark_output_directory', and 'non_blurry_output_directory' as needed
\ No newline at end of file

similaritycheck.py

+import os
+import cv2
+import shutil
+from PIL import Image
+from imagehash import dhash
+
+# Define the threshold for similarity
+similarity_threshold = 10  # You can adjust this value based on your needs
+
+# Get the directory where the script is located
+script_directory = os.path.dirname(__file__)
+
+# Create a directory to store similar images
+similar_output_directory = os.path.join(script_directory, 'similar_images')
+
+# Create a directory to store unique images
+unique_output_directory = os.path.join(script_directory, 'unique_images')
+
+# Function to empty a folder
+def empty_folder(folder):
+    for filename in os.listdir(folder):
+        file_path = os.path.join(folder, filename)
+        try:
+            if os.path.isfile(file_path):
+                os.unlink(file_path)
+        except Exception as e:
+            print(e)
+
+# Empty the similar and unique image folders if they are non-empty
+if os.path.exists(similar_output_directory):
+    empty_folder(similar_output_directory)
+if os.path.exists(unique_output_directory):
+    empty_folder(unique_output_directory)
+
+# Dictionary to store image hashes and their paths
+image_hashes = {}
+
+# Use os.walk to traverse all subdirectories
+for folder, subfolders, files in os.walk(script_directory):
+    for file in files:
+        # Check if the file is an image (you can customize the extensions)
+        if file.lower().endswith(('.jpg', '.jpeg', '.png', '.gif', '.bmp')):
+            image_path = os.path.join(folder, file)
+            
+            # Load the image
+            image = cv2.imread(image_path)
+            
+            # Convert the NumPy array to a PIL image object
+            pil_image = Image.fromarray(cv2.cvtColor(image, cv2.COLOR_BGR2RGB))
+            
+            # Calculate the dHash for the image
+            image_hash = dhash(pil_image, hash_size=8)
+            
+            # Check if this hash is similar to any previously seen hashes
+            similar = False
+            for existing_hash, existing_path in image_hashes.items():
+                hamming_distance = image_hash - existing_hash
+                if hamming_distance < similarity_threshold:
+                    similar = True
+                    print(f"Similar Image: {image_path} is similar to {existing_path}")
+                    if not os.path.exists(similar_output_directory):
+                        os.makedirs(similar_output_directory)
+                    destination_path = os.path.join(similar_output_directory, os.path.basename(image_path))
+                    shutil.copy2(image_path, destination_path)
+                    break
+
+            # If not similar to any previously seen images, add it to the dictionary and copy it to the unique folder
+            if not similar:
+                image_hashes[image_hash] = image_path
+                print(f"Unique Image: {image_path}")
+                if not os.path.exists(unique_output_directory):
+                    os.makedirs(unique_output_directory)
+                destination_path = os.path.join(unique_output_directory, os.path.basename(image_path))
+                shutil.copy2(image_path, destination_path)
+
+# You can adjust the 'similarity_threshold', 'similar_output_directory', and 'unique_output_directory' as needed
\ No newline at end of file