image_to_text / app.py
amoldwalunj's picture
Update app.py
adec58c
import pytesseract
from PIL import Image
import pandas as pd
import os
import gradio as gr
import numpy as np
import gradio as gr
def ocr_df_using_pytesseract(image):
#pytesseract.pytesseract.tesseract_cmd =r"C:\Users\amold\Desktop\Upwork\pdf to image and pytesseract\tesseact_exe\Tesseract-OCR\tesseract.exe"
pytesseract.pytesseract.tesseract_cmd = r'/usr/bin/tesseract'
#image = Image.open(example['image_path'])
width, height = image.size
# apply ocr to the image
ocr_df = pytesseract.image_to_data(image, output_type='data.frame')
float_cols = ocr_df.select_dtypes('float').columns
ocr_df = ocr_df.dropna().reset_index(drop=True)
ocr_df[float_cols] = ocr_df[float_cols].round(0).astype(int)
ocr_df = ocr_df.replace(r'^\s*$', np.nan, regex=True)
ocr_df = ocr_df.dropna().reset_index(drop=True)
ocr_df
ocr_df['X1']=ocr_df['left']
ocr_df['Y1']=ocr_df['top']
ocr_df['X2']= ocr_df['left'] + ocr_df['width']
ocr_df['Y2']= ocr_df['top'] + ocr_df['height']
return ocr_df
def image_to_text(image):
ocr_df= ocr_df_using_pytesseract(image)
grouped_text = ocr_df.groupby(['block_num', 'line_num'])['text'].agg(' '.join).reset_index()
# sort the text by line numbers within each block
grouped_text = grouped_text.sort_values(['block_num', 'line_num'])
# join the text by blocks and add newlines
result = ''
for i, row in grouped_text.iterrows():
if i > 0 and row['block_num'] != grouped_text.loc[i-1, 'block_num']:
result += '\n\n'
result += row['text'].rstrip() + '\n'
return result
demo = gr.Interface(fn=image_to_text,
inputs= gr.Image(type="pil"),
outputs=["text"],
title="Menu image to text",
description= "Upload Menu image here")
demo.launch(share=False)