Quickly annotate cropped text images for training your OCR-model

quick_and_easy_annotator.py

import os
from tkinter import *
from tkinter import ttk
import sys
import pandas as pd
from PIL import ImageTk,Image

base = Tk()
base.geometry("500x500")
base.resizable(width = True, height = True)

LARGE_FONT= ("Verdana", 12)
NORM_FONT= ("Verdana", 10)
SMALL_FONT= ("Verdana", 8)

def popupmsg(msg):
    popup = Tk()
    popup.wm_title("!")
    label = ttk.Label(popup, text=msg, font=NORM_FONT)
    label.pack(side="top", fill="x", pady=10)
    B1 = ttk.Button(popup, text="Okay", command = popup.destroy)
    B1.pack()
    popup.mainloop()

def load_image():
    global base, img, image_names,image_dir, images_dir,image_name,cimg,box_entry
    try:
        image_name = image_names.pop()
    except:
        popupmsg('No images Left to Annotate!')
        pass
    box_entry.delete(0,END)
    image_dir = os.path.join(images_dir,image_name)
    img = ImageTk.PhotoImage(Image.open(image_dir))
    canvas.delete("all")
    canvas.create_image(250,250,image=img,anchor=CENTER)

def save_data():
    global base, img, image_names, image_dir, images_dir, cimg, box_entry
    data = box_entry.get()
    with open(images_dir+".txt","a") as f:
        f.write(image_name+","+data+"\n")
    load_image()

def on_enter_key(event):
    save_data()

def initialize_file():
    global image_names
    if os.path.exists(images_dir+'.txt'):
# For removing duplicates in the existing csv file
        labels_file = pd.read_csv(images_dir+'.txt')
        labels_file.columns = ["image name","label"]
        labels_file = labels_file.drop_duplicates(keep='first')
        labels_file = labels_file.drop_duplicates(subset="image name",keep=False)
        labels_file.to_csv(images_dir+'.txt',columns=["image name","label"],index=False)
# No need to do files which have already been done
        col1 = labels_file["image name"].unique()
        for imagei in image_names[:]:
            if imagei in col1:
                image_names.remove(imagei)

submit_button = Button(base,text = 'Submit', command = save_data).grid(row = 0, column = 0)
skip_button = Button(base,text = 'Skip',command = load_image).grid(row = 0, column = 1)
annotation_label = Label(base, text = 'Annotation').grid(row = 1, column = 0)
box_entry = Entry(base)
box_entry.grid(row=1,column=1)

if len(sys.argv)<2:
    images_dir = 'saved'
else:
    images_dir = sys.argv[1]

image_names = os.listdir(images_dir)
image_names.sort()

initialize_file()
#print(image_names)
canvas = Canvas(base, width = 500, height = 500)
try:
    image_name = image_names.pop()
except:
    print("No images left to annotate in the specified folder")
    exit(0)
    popupmsg("No images left to annotate in the specified folder!")
image_dir = os.path.join(images_dir,image_name)
img = ImageTk.PhotoImage(Image.open(image_dir))
cimg = canvas.create_image(250,250,image = img,anchor=CENTER)
canvas.grid(row=2,columnspan=2)
box_entry.bind('<Return>',on_enter_key)
box_entry.focus()
base.mainloop()