Compare OCR results from images
Display OCRBench leaderboard for model evaluations
Convert images of text into editable text
Qwen2-VL is a vision-language model that performs OCR
Python3 package for Chinese/English OCR, with paddleocr-v4 o
Upload an image to extract, correct, and spell-check text
Surya OCR
Extract text from images using OCR
Surya OCR
NepaliOCR
Consist of HOG LR, CRNN, and TrOCR
Extract text and tables from French images
Identify lottery numbers from images
Streamlit OCR Comparator is a web-based application built using Streamlit that enables users to compare OCR (Optical Character Recognition) results from different OCR engines. It provides a user-friendly interface to upload images, extract text using multiple OCR tools, and visually compare the results. This tool is particularly useful for developers and researchers to evaluate and benchmark OCR performance across various engines.
• Multi-OCR Engine Support: Compare text extraction results from multiple OCR engines in a single interface.
• Image Upload Capability: Upload images in popular formats (JPEG, PNG, etc.) for OCR processing.
• Real-Time Comparison: Get instantaneous results from different OCR engines for quick evaluation.
• Accuracy Metrics: View accuracy scores and visual differences between OCR outputs.
• Side-by-Side Display: Easily compare OCR results from different engines side-by-side.
pip install streamlit-ocr-comparator in your terminal to install the package.streamlit run ocr_comparator.py to start the application.What OCR engines are supported by Streamlit OCR Comparator?
Streamlit OCR Comparator supports popular OCR engines like Tesseract, Google Cloud Vision API, and Microsoft Azure Computer Vision.
How do I upload multiple images at once?
Currently, the application only supports single image uploads. You can process multiple images sequentially.
Can I save the comparison results?
Yes, you can download the comparison results as a text file or CSV for further analysis.