Commit
·
3ecbba0
1
Parent(s):
df33ba6
fix: resolve merge conflicts and update UI
Browse files
app.py
CHANGED
|
@@ -11,18 +11,22 @@ import logging
|
|
| 11 |
import cv2
|
| 12 |
from datetime import datetime
|
| 13 |
import time
|
|
|
|
| 14 |
|
| 15 |
# Set up logging for error handling
|
| 16 |
logging.basicConfig(level=logging.DEBUG)
|
|
|
|
| 17 |
|
| 18 |
# Initialize the OCR reader
|
| 19 |
reader = easyocr.Reader(['en'])
|
| 20 |
|
| 21 |
-
#
|
| 22 |
-
|
| 23 |
-
|
| 24 |
-
|
| 25 |
-
|
|
|
|
|
|
|
| 26 |
|
| 27 |
def convert_pdf_to_images(pdf_file):
|
| 28 |
"""Convert PDF to list of images with detailed logging"""
|
|
@@ -96,93 +100,96 @@ def process_single_image(image):
|
|
| 96 |
logging.error(f"Error processing the image: {str(e)}")
|
| 97 |
return {'error': f"Error processing the image: {str(e)}"}
|
| 98 |
|
| 99 |
-
|
| 100 |
-
|
| 101 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 102 |
try:
|
|
|
|
|
|
|
|
|
|
|
|
|
| 103 |
if input_file.name.lower().endswith('.pdf'):
|
| 104 |
images = convert_pdf_to_images(input_file)
|
| 105 |
if images is None:
|
| 106 |
-
return
|
| 107 |
result = process_single_image(images[0])
|
| 108 |
-
elif input_file.name.lower().endswith(('png', 'jpg', 'jpeg')):
|
| 109 |
img = Image.open(input_file)
|
| 110 |
result = process_single_image(img)
|
| 111 |
else:
|
| 112 |
-
return
|
|
|
|
|
|
|
|
|
|
| 113 |
|
| 114 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 115 |
|
| 116 |
except Exception as e:
|
| 117 |
-
|
| 118 |
-
return
|
| 119 |
|
| 120 |
-
# Create the Gradio interface
|
| 121 |
iface = gr.Interface(
|
| 122 |
-
fn=
|
| 123 |
inputs=[
|
| 124 |
gr.File(
|
| 125 |
label="Upload Insurance Document",
|
| 126 |
-
file_types=["pdf", "png", "jpg", "jpeg", "tiff"]
|
| 127 |
-
type="filepath"
|
| 128 |
)
|
| 129 |
],
|
| 130 |
outputs=[
|
| 131 |
-
gr.Textbox(
|
| 132 |
-
|
| 133 |
-
|
| 134 |
-
|
| 135 |
-
),
|
| 136 |
-
gr.Textbox(
|
| 137 |
-
label="Validation Results",
|
| 138 |
-
lines=3,
|
| 139 |
-
show_copy_button=True
|
| 140 |
-
),
|
| 141 |
-
gr.Textbox(
|
| 142 |
-
label="Document Classification",
|
| 143 |
-
lines=3,
|
| 144 |
-
show_copy_button=True
|
| 145 |
-
),
|
| 146 |
-
gr.File(
|
| 147 |
-
label="Download Analysis Report (Excel)",
|
| 148 |
-
type="filepath"
|
| 149 |
-
)
|
| 150 |
],
|
| 151 |
title="Insurance Claim Document Analyzer",
|
| 152 |
description="Upload insurance documents (PDF/Images) for automated text extraction, validation, and classification.",
|
| 153 |
-
|
| 154 |
-
|
| 155 |
-
- Multi-page PDF processing
|
| 156 |
-
- Image formats: PNG, JPG, TIFF
|
| 157 |
-
- Text extraction & validation
|
| 158 |
-
- Document classification
|
| 159 |
-
- Detailed Excel report generation
|
| 160 |
-
|
| 161 |
-
### File Requirements:
|
| 162 |
-
- Maximum file size: 10MB
|
| 163 |
-
- Clear, readable content
|
| 164 |
-
- English language documents
|
| 165 |
-
""",
|
| 166 |
-
examples=[],
|
| 167 |
-
cache_examples=False,
|
| 168 |
-
theme=gr.themes.Soft(
|
| 169 |
-
primary_hue="blue",
|
| 170 |
-
secondary_hue="gray",
|
| 171 |
-
neutral_hue="gray"
|
| 172 |
-
),
|
| 173 |
-
css=".gradio-container {max-width: 900px; margin: auto}",
|
| 174 |
-
allow_flagging="never",
|
| 175 |
-
analytics_enabled=False
|
| 176 |
)
|
| 177 |
|
| 178 |
-
# Launch with
|
| 179 |
if __name__ == "__main__":
|
| 180 |
-
|
| 181 |
-
|
| 182 |
-
|
| 183 |
-
|
| 184 |
-
|
| 185 |
-
|
| 186 |
-
|
| 187 |
-
|
| 188 |
-
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
import cv2
|
| 12 |
from datetime import datetime
|
| 13 |
import time
|
| 14 |
+
import os
|
| 15 |
|
| 16 |
# Set up logging for error handling
|
| 17 |
logging.basicConfig(level=logging.DEBUG)
|
| 18 |
+
logger = logging.getLogger(__name__)
|
| 19 |
|
| 20 |
# Initialize the OCR reader
|
| 21 |
reader = easyocr.Reader(['en'])
|
| 22 |
|
| 23 |
+
# Initialize models with error handling
|
| 24 |
+
try:
|
| 25 |
+
text_classifier = pipeline("text-classification", model="distilbert-base-uncased-finetuned-sst-2-english")
|
| 26 |
+
doc_classifier = pipeline("image-classification", model="microsoft/resnet-50")
|
| 27 |
+
except Exception as e:
|
| 28 |
+
logger.error(f"Error initializing models: {str(e)}")
|
| 29 |
+
raise
|
| 30 |
|
| 31 |
def convert_pdf_to_images(pdf_file):
|
| 32 |
"""Convert PDF to list of images with detailed logging"""
|
|
|
|
| 100 |
logging.error(f"Error processing the image: {str(e)}")
|
| 101 |
return {'error': f"Error processing the image: {str(e)}"}
|
| 102 |
|
| 103 |
+
def generate_excel_report(result):
|
| 104 |
+
"""Generate Excel report from processing results"""
|
| 105 |
+
try:
|
| 106 |
+
df = pd.DataFrame([{
|
| 107 |
+
'Timestamp': datetime.now().strftime("%Y-%m-%d %H:%M:%S"),
|
| 108 |
+
'Extracted Text': result.get('text', ''),
|
| 109 |
+
'Validation': result.get('validation', ''),
|
| 110 |
+
'Validation Confidence': result.get('validation_confidence', 0),
|
| 111 |
+
'Document Type': result.get('doc_type', ''),
|
| 112 |
+
'Document Confidence': result.get('doc_confidence', 0)
|
| 113 |
+
}])
|
| 114 |
+
|
| 115 |
+
output_path = f"report_{datetime.now().strftime('%Y%m%d_%H%M%S')}.xlsx"
|
| 116 |
+
df.to_excel(output_path, index=False)
|
| 117 |
+
return output_path
|
| 118 |
+
except Exception as e:
|
| 119 |
+
logger.error(f"Error generating report: {str(e)}")
|
| 120 |
+
return None
|
| 121 |
+
|
| 122 |
+
def process_claim(input_file):
|
| 123 |
+
"""Main processing function for the Gradio interface"""
|
| 124 |
try:
|
| 125 |
+
if input_file is None:
|
| 126 |
+
return "Please upload a file.", "No file provided.", "No file provided.", None
|
| 127 |
+
|
| 128 |
+
# Process the file
|
| 129 |
if input_file.name.lower().endswith('.pdf'):
|
| 130 |
images = convert_pdf_to_images(input_file)
|
| 131 |
if images is None:
|
| 132 |
+
return "Error processing PDF.", "Invalid PDF file.", "Processing failed.", None
|
| 133 |
result = process_single_image(images[0])
|
| 134 |
+
elif input_file.name.lower().endswith(('png', 'jpg', 'jpeg', 'tiff')):
|
| 135 |
img = Image.open(input_file)
|
| 136 |
result = process_single_image(img)
|
| 137 |
else:
|
| 138 |
+
return "Unsupported file type.", "Invalid file format.", "Processing failed.", None
|
| 139 |
+
|
| 140 |
+
if 'error' in result:
|
| 141 |
+
return result['error'], "Processing failed.", "Processing failed.", None
|
| 142 |
|
| 143 |
+
# Generate Excel report
|
| 144 |
+
report_path = generate_excel_report(result)
|
| 145 |
+
|
| 146 |
+
# Format output strings
|
| 147 |
+
validation_text = f"Validation: {result['validation']}\nConfidence: {result['validation_confidence']:.2%}"
|
| 148 |
+
classification_text = f"Type: {result['doc_type']}\nConfidence: {result['doc_confidence']:.2%}"
|
| 149 |
+
|
| 150 |
+
return (
|
| 151 |
+
result['text'],
|
| 152 |
+
validation_text,
|
| 153 |
+
classification_text,
|
| 154 |
+
report_path if report_path else None
|
| 155 |
+
)
|
| 156 |
|
| 157 |
except Exception as e:
|
| 158 |
+
logger.error(f"Error in process_claim: {str(e)}")
|
| 159 |
+
return str(e), "Processing failed.", "Processing failed.", None
|
| 160 |
|
| 161 |
+
# Create the Gradio interface
|
| 162 |
iface = gr.Interface(
|
| 163 |
+
fn=process_claim,
|
| 164 |
inputs=[
|
| 165 |
gr.File(
|
| 166 |
label="Upload Insurance Document",
|
| 167 |
+
file_types=["pdf", "png", "jpg", "jpeg", "tiff"]
|
|
|
|
| 168 |
)
|
| 169 |
],
|
| 170 |
outputs=[
|
| 171 |
+
gr.Textbox(label="Extracted Text", lines=10),
|
| 172 |
+
gr.Textbox(label="Validation Results", lines=3),
|
| 173 |
+
gr.Textbox(label="Document Classification", lines=3),
|
| 174 |
+
gr.File(label="Download Analysis Report")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 175 |
],
|
| 176 |
title="Insurance Claim Document Analyzer",
|
| 177 |
description="Upload insurance documents (PDF/Images) for automated text extraction, validation, and classification.",
|
| 178 |
+
theme=gr.themes.Soft(primary_hue="blue"),
|
| 179 |
+
css=".gradio-container {max-width: 900px; margin: auto}"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 180 |
)
|
| 181 |
|
| 182 |
+
# Launch with error handling
|
| 183 |
if __name__ == "__main__":
|
| 184 |
+
try:
|
| 185 |
+
iface.launch(
|
| 186 |
+
server_name="0.0.0.0",
|
| 187 |
+
server_port=7860,
|
| 188 |
+
share=False,
|
| 189 |
+
debug=True,
|
| 190 |
+
enable_queue=True,
|
| 191 |
+
max_threads=4
|
| 192 |
+
)
|
| 193 |
+
except Exception as e:
|
| 194 |
+
logger.error(f"Failed to launch interface: {str(e)}")
|
| 195 |
+
raise
|