gui_arayüz

#oluşturulan tkinter arayüzünün kodları:

import tkinter as tk
from tkinter import ttk
from PIL import ImageTk, Image
from tkinter import filedialog
import numpy as np
import tensorflow
from tensorflow.keras.preprocessing.image import img_to_array
from tensorflow.keras.preprocessing import image

CLASS_LABELS = ['Kızgın',
                'İğrenmiş',
                'Korkmuş',
                'Mutlu',
                'Nötr',
                'Üzgün',
                "Şaşkın",
                ]

COLORS = ['#9b5de5',
          '#f15bb5',
          '#f8a07b',
          '#fee440',
          '#7fd09d',
          '#00bbf9',
          '#00f5d4',
          ]


def decode_predictions(predictions):
    return sorted(list(zip(CLASS_LABELS, predictions[0])), key=lambda x: x[1], reverse=True)


def load_img():

    global img, image_data
    for img_display in frame.winfo_children():
        img_display.destroy()

    image_data = filedialog.askopenfilename(initialdir="/", title="Choose an image",
                                            filetypes=(("all files", "*.*"), ("png files", "*.png")))
    basewidth = 150
    img = Image.open(image_data)
    wpercent = (basewidth / float(img.size[0]))
    hsize = int((float(img.size[1]) * float(wpercent)))
    img = img.resize((basewidth, hsize), Image.ANTIALIAS)
    img = ImageTk.PhotoImage(img)
    file_name = image_data.split('/')

    panel = tk.Label(frame,
                     text=str(
                         file_name[-1]),
                     width=150)
    # panel.pack()
    panel.grid(row=0, column=0, columnspan=2)

    panel_image = tk.Label(frame, image=img)
    # panel_image.pack()
    panel_image.grid(row=1, column=0, columnspan=2)

    classify()


def classify():

    # Preprocessing
    original = Image.open(image_data).convert('L')
    original = original.resize((48, 48), Image.ANTIALIAS)
    img_rgb = original.convert('RGB')
    img_arr = image.img_to_array(img_rgb)
    img_batch = np.expand_dims(img_arr, axis=0)
    img_preprocessed = img_batch.astype(np.float32) / 255.0

    # Predicting
    predictions = model.predict(img_preprocessed)

    label = decode_predictions(predictions)
    table = tk.Label(
        frame, text="Tahminler", font="bold", bg='#181818', fg='#f9f6ee')
    # table.pack()
    table.grid(row=2, column=0, columnspan=2)

    for i in range(0, len(label)):
        result_label = tk.Label(frame,
                                text=str(label[i][0]),
                                bg="#181818",
                                fg=COLORS[i],  # "#f9f6ee",
                                font="bold",
                                )
        result_prediction = tk.Label(frame,
                                     text=str(
                                         round(float(label[i][1])*100, 2)) + '%',
                                     bg="#181818",
                                     fg=COLORS[i],  # "#f9f6ee",
                                     font="bold",
                                     )
        result_label.grid(row=i+3, column=0, sticky='ew')
        result_prediction.grid(row=i+3, column=1, sticky='ew')

        frame.grid_columnconfigure((0, 1), weight=1)
        # result.pack()


root = tk.Tk()
root.title('IMAGE CLASSIFIER')
root.configure(background='#181818')
root.resizable(False, False)

canvas = tk.Canvas(root, height=600, width=600,
                   bg='#181818', highlightthickness=0)
canvas.grid(row=0, column=0)
# canvas.pack()

frame = tk.Frame(root,
                 bg='#181818',
                 highlightbackground='#f9f6ee',
                 highlightthickness=3)  # 2e74b7
frame.place(relwidth=0.8, relheight=0.8, relx=0.1, rely=0.1)

chose_image = tk.Button(root, text='Choose Image',
                        padx=35, pady=10, font='bold', borderwidth=0,
                        fg="#f9f6ee", bg="#181818", command=load_img,
                        )
chose_image.grid(row=1, column=0)

model = tensorflow.keras.models.load_model('checkpoint3.h5')

root.mainloop()