Image Duplicate Finder (Hash-Based)

import os

import tkinter as tk

from tkinter import filedialog, ttk, messagebox

from PIL import Image, ImageTk

import imagehash



class ImageDuplicateFinder:

    def __init__(self, root):

        self.root = root

        self.root.title("Image Duplicate Finder (Hash-Based)")

        self.root.geometry("900x600")


        self.image_hash_map = {}

        self.duplicates = []


        self.setup_ui()


    # ---------------------------------------------------------

    def setup_ui(self):

        top_frame = tk.Frame(self.root)

        top_frame.pack(fill=tk.X, pady=10)


        tk.Button(top_frame, text="Select Folder", command=self.select_folder).pack(side=tk.LEFT, padx=10)

        tk.Button(top_frame, text="Scan for Duplicates", command=self.scan_duplicates).pack(side=tk.LEFT, padx=10)


        self.result_tree = ttk.Treeview(self.root, columns=("img1", "img2"), show="headings")

        self.result_tree.heading("img1", text="Image 1")

        self.result_tree.heading("img2", text="Duplicate Image")

        self.result_tree.pack(fill="both", expand=True, pady=10)


        self.result_tree.bind("<Double-1>", self.show_preview)


    # ---------------------------------------------------------

    def select_folder(self):

        self.folder_path = filedialog.askdirectory()

        if self.folder_path:

            messagebox.showinfo("Selected Folder", self.folder_path)


    # ---------------------------------------------------------

    def scan_duplicates(self):

        if not hasattr(self, "folder_path"):

            messagebox.showerror("Error", "Please select a folder first!")

            return


        self.image_hash_map = {}

        self.duplicates = []

        self.result_tree.delete(*self.result_tree.get_children())


        supported_ext = (".jpg", ".jpeg", ".png", ".bmp")


        for filename in os.listdir(self.folder_path):

            if filename.lower().endswith(supported_ext):

                full_path = os.path.join(self.folder_path, filename)


                try:

                    img = Image.open(full_path)

                    img_hash = imagehash.average_hash(img)


                    if img_hash in self.image_hash_map:

                        self.duplicates.append((self.image_hash_map[img_hash], full_path))


                    else:

                        self.image_hash_map[img_hash] = full_path


                except Exception as e:

                    print("Error loading:", filename, e)


        # Show results in table

        for dup in self.duplicates:

            self.result_tree.insert("", tk.END, values=dup)


        if not self.duplicates:

            messagebox.showinfo("Result", "No duplicates found!")

        else:

            messagebox.showinfo("Result", f"Found {len(self.duplicates)} duplicate pairs.")


    # ---------------------------------------------------------

    def show_preview(self, event):

        selected = self.result_tree.focus()

        if not selected:

            return


        img1_path, img2_path = self.result_tree.item(selected, "values")

        self.preview_window(img1_path, img2_path)


    # ---------------------------------------------------------

    def preview_window(self, img1_path, img2_path):

        win = tk.Toplevel(self.root)

        win.title("Duplicate Image Preview")

        win.geometry("850x450")


        tk.Label(win, text=f"Image 1: {os.path.basename(img1_path)}", font=("Arial", 10)).pack(pady=5)

        tk.Label(win, text=f"Image 2: {os.path.basename(img2_path)}", font=("Arial", 10)).pack(pady=5)


        frame = tk.Frame(win)

        frame.pack(expand=True)


        # Load images

        img1 = Image.open(img1_path).resize((350, 350))

        img2 = Image.open(img2_path).resize((350, 350))


        img1_tk = ImageTk.PhotoImage(img1)

        img2_tk = ImageTk.PhotoImage(img2)


        left_label = tk.Label(frame, image=img1_tk)

        left_label.image = img1_tk

        left_label.pack(side=tk.LEFT, padx=10)


        right_label = tk.Label(frame, image=img2_tk)

        right_label.image = img2_tk

        right_label.pack(side=tk.RIGHT, padx=10)



# ---------------------------------------------------------

# RUN APP

# ---------------------------------------------------------

root = tk.Tk()

app = ImageDuplicateFinder(root)

root.mainloop()

 

No comments: