first version
Browse files
app.py
ADDED
|
@@ -0,0 +1,53 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gr
|
| 2 |
+
import paddle
|
| 3 |
+
from paddleocr import PaddleOCR, draw_ocr
|
| 4 |
+
import cv2
|
| 5 |
+
from PIL import Image
|
| 6 |
+
import numpy as np
|
| 7 |
+
|
| 8 |
+
# Initialize PaddleOCR for Arabic
|
| 9 |
+
ocr = PaddleOCR(use_angle_cls=True, lang='ar')
|
| 10 |
+
|
| 11 |
+
def ocr_extract_text(image):
|
| 12 |
+
# Perform OCR on the uploaded image
|
| 13 |
+
result = ocr.ocr(image, cls=True)
|
| 14 |
+
|
| 15 |
+
# Initialize a list to store detected text
|
| 16 |
+
detected_text = []
|
| 17 |
+
|
| 18 |
+
# Loop through the results and extract text
|
| 19 |
+
for line in result[0]: # result[0] contains the OCR results for text regions
|
| 20 |
+
text = line[1][0] # Detected text
|
| 21 |
+
detected_text.append(text)
|
| 22 |
+
|
| 23 |
+
# Draw the OCR results on the image
|
| 24 |
+
boxes = [line[0] for line in result[0]] # Bounding box coordinates
|
| 25 |
+
texts = [line[1][0] for line in result[0]] # Detected text
|
| 26 |
+
scores = [line[1][1] for line in result[0]] # Confidence scores
|
| 27 |
+
|
| 28 |
+
annotated_image = draw_ocr(image, boxes, texts, scores, font_path="path/to/arabic/font.ttf")
|
| 29 |
+
|
| 30 |
+
# Convert the annotated image to a format that Gradio can display
|
| 31 |
+
annotated_image = Image.fromarray(np.uint8(annotated_image))
|
| 32 |
+
|
| 33 |
+
# Return the text and annotated image
|
| 34 |
+
return "\n".join(detected_text), annotated_image
|
| 35 |
+
|
| 36 |
+
# Define Gradio interface with the button to trigger OCR
|
| 37 |
+
def start_ocr(image):
|
| 38 |
+
# Trigger OCR extraction when button is clicked
|
| 39 |
+
return ocr_extract_text(image)
|
| 40 |
+
|
| 41 |
+
iface = gr.Interface(
|
| 42 |
+
fn=start_ocr, # Function to trigger OCR extraction on button click
|
| 43 |
+
inputs=gr.Image(type="pil"), # Image input only
|
| 44 |
+
outputs=[gr.Textbox(label="Extracted Text"), gr.Image(label="Annotated Image")], # Outputs
|
| 45 |
+
live=False, # Set live=False as we want to trigger the process with a button click
|
| 46 |
+
title="Arabic OCR Extractor", # Title of the interface
|
| 47 |
+
description="Upload an Arabic document or image, and click 'Start Extracting Text' to extract the text using OCR.", # Description
|
| 48 |
+
allow_flagging="never" # Prevent flagging if it's not required
|
| 49 |
+
)
|
| 50 |
+
|
| 51 |
+
if __name__ == "__main__":
|
| 52 |
+
# This ensures Gradio runs in debug mode and with queue for better debugging
|
| 53 |
+
iface.queue().launch(debug=True)
|