Unofficial demo for TB-OCR (OCR for documents)
Extract text from PDFs
Convert handwritten images to text
Demo of GOT-OCR 2.0's Transformers implementation
Extract text from images using OCR
Give it a pdf and it'll extract the text
Extract text from vehicle number plates
Extract text from images using OCR
Scan and extract text from documents
Convert images to multiplication pairs text
Extract text from images using OCR
Extract text from Tifinagh images
Qwen2-VL is a vision-language model that performs OCR
TB-OCR Demo is an unofficial demo application designed to demonstrate the capabilities of TB-OCR, an Optical Character Recognition (OCR) tool. It allows users to extract and convert text from images of documents into editable markdown text. This tool is particularly useful for digitizing handwritten notes, scanned documents, and other image-based text content, making it accessible for further editing or sharing.
What file formats does TB-OCR Demo support?
TB-OCR Demo supports common image formats such as JPEG, PNG, and BMP for input. The output is provided in markdown text format.
Can TB-OCR Demo handle handwritten text?
Yes, TB-OCR Demo includes handwriting support, allowing it to recognize and convert handwritten text from images, though accuracy may vary depending on handwriting quality.
How do I handle complex document layouts?
For complex layouts, you can pre-process the image by cropping specific sections or using the built-in options to improve OCR accuracy. The results can then be manually adjusted if needed.