Spaces:
Sleeping
Sleeping
File size: 1,127 Bytes
513f893 65ef5f8 d22d28e 65ef5f8 d22d28e 513f893 65ef5f8 513f893 d4534d1 65ef5f8 d22d28e 513f893 af7cef1 d22d28e d4534d1 d22d28e d4534d1 af7cef1 d22d28e 1cb8b90 d22d28e af7cef1 d4534d1 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 |
from transformers import TrOCRProcessor, VisionEncoderDecoderModel
from PIL import Image
import re
# Load model + processor
processor = TrOCRProcessor.from_pretrained("microsoft/trocr-base-stage1")
model = VisionEncoderDecoderModel.from_pretrained("microsoft/trocr-base-stage1")
def extract_weight(image: Image.Image) -> str:
image = image.convert("RGB")
pixel_values = processor(images=image, return_tensors="pt").pixel_values
generated_ids = model.generate(pixel_values)
full_text = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
# Lowercase text but don't strip spacing before kg detection
full_text_lower = full_text.lower()
# Detect unit
if "kg" in full_text_lower.replace(" ", ""):
unit = "kg"
elif "g" in full_text_lower.replace(" ", "") or "gram" in full_text_lower:
unit = "grams"
else:
unit = "grams" # default
# Extract number using regex
match = re.search(r"(\d+(\.\d+)?)", full_text)
if match:
weight = match.group(1)
return f"{weight} {unit}"
else:
return "No valid weight detected"
|