from typing import List import pytesseract from PIL import Image import gradio as gr import cv2 import numpy as np def tesseract_ocr_with_selection(filepath: str, coordinates: List[int] = None): # Görseli yükle image = Image.open(filepath) if coordinates: # Koordinatlara göre kırp x1, y1, x2, y2 = coordinates image = image.crop((x1, y1, x2, y2)) # OCR işlemi (varsayılan dil: İngilizce) return pytesseract.image_to_string(image=image, lang='eng') def parse_coordinates(coord_input: str): """ Kullanıcıdan alınan koordinat stringini doğrula ve liste olarak döndür. """ try: # Koordinatları virgül ile ayır ve tam sayıya çevir coords = [int(coord.strip()) for coord in coord_input.split(",")] if len(coords) != 4: raise ValueError("Lütfen tam olarak 4 koordinat girin (örnek: x1, y1, x2, y2).") return coords except ValueError: raise ValueError("Hatalı koordinat formatı. Lütfen şu formatı kullanın: x1, y1, x2, y2.") # Gradio UI ayarları title = "Tesseract OCR with Selection" description = "Gradio demo for Tesseract OCR with region selection (default language: English)." article = "

Tesseract documentation | Github Repo

" # examples = [ # ['examples/eurotext.png', "50, 50, 200, 200"], # ['examples/tesseract_sample.png', "30, 40, 150, 120"], # ] with gr.Blocks() as demo: with gr.Row(): gr.Markdown("# Tesseract OCR with Selection") with gr.Row(): img_input = gr.Image(type="filepath", label="Input Image") coords_input = gr.Textbox(label="Selection Coordinates (x1, y1, x2, y2)", placeholder="50, 50, 200, 200") with gr.Row(): ocr_button = gr.Button("Run OCR with Selection") with gr.Row(): ocr_output = gr.Textbox(label="OCR Result") def run_with_selection(image_path, coordinates): try: # Koordinatları doğrula ve ayrıştır coords = parse_coordinates(coordinates) return tesseract_ocr_with_selection(image_path, coords) except ValueError as e: return str(e) # Kullanıcıya hata mesajı göster ocr_button.click( run_with_selection, inputs=[img_input, coords_input], outputs=[ocr_output] ) if __name__ == '__main__': demo.launch()