Spaces:
Sleeping
Sleeping
Upload ocr_processor.py
Browse files- src/ocr_processor.py +23 -0
src/ocr_processor.py
ADDED
|
@@ -0,0 +1,23 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import easyocr
|
| 2 |
+
from text_cleaner import clean_product_text, extract_keywords
|
| 3 |
+
import re
|
| 4 |
+
|
| 5 |
+
def extract_info(image_path):
|
| 6 |
+
reader = easyocr.Reader(['en'])
|
| 7 |
+
results = reader.readtext(image_path, detail=0)
|
| 8 |
+
raw_text = " ".join(results)
|
| 9 |
+
|
| 10 |
+
# Clean and extract the info we did
|
| 11 |
+
clean_text = clean_product_text(raw_text)
|
| 12 |
+
brand, product_name = extract_keywords(clean_text)
|
| 13 |
+
|
| 14 |
+
# Barcode detection
|
| 15 |
+
barcodes = re.findall(r'\b(\d{12,13})\b', raw_text)
|
| 16 |
+
|
| 17 |
+
return {
|
| 18 |
+
"raw_text": raw_text,
|
| 19 |
+
"clean_text": clean_text,
|
| 20 |
+
"brand": brand,
|
| 21 |
+
"product_name": product_name,
|
| 22 |
+
"barcodes": barcodes
|
| 23 |
+
}
|