How to get or see the re-constructed page for a given input ?

[PrasannaVpk]

I want to see the result of page-constructed image,how to get that ?

[fg-mindee]

Hi @PrasannaVpk 👋

If you mean the image synthesized from the model predictions, here is a snippet:

import matplotlib.pyplot as plt
from doctr.io import DocumentFile
from doctr.models import ocr_predictor
from doctr.utils.visualization import synthetize_page

model = ocr_predictor(pretrained=True)
# PDF
doc = DocumentFile.from_pdf("path/to/your/doc.pdf").as_images()
# Analyze
result = model(doc)
# Reconstructed the first page of the document
img = synthetize_page(result.pages[0].export())
plt.imshow(img); plt.show()

which yields:

Hope this helps :)
If I misunderstood you, please specify what you meant so that we can assist you 👌

[PrasannaVpk]

Yes,this is what i expected . Thanks for the quick response.