Instructions to use foduucom/table-detection-and-extraction with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- ultralytics
How to use foduucom/table-detection-and-extraction with ultralytics:
from ultralytics import YOLOvv8 model = YOLOvv8.from_pretrained("foduucom/table-detection-and-extraction") source = 'http://images.cocodataset.org/val2017/000000039769.jpg' model.predict(source=source, save=True) - Notebooks
- Google Colab
- Kaggle
I am having multiple tables and want only a specified one how can I achieve, and also how to extract the table
I am having around 3-4 tables and i want only a specified table how can i classify that. Like if i want the table which is bigger is that the right approach. Later on how can i extract the data from the table.
Instead of finding the larger table,. I would suggest looking into the table data, which looks relevant to you after performing the OCR.
As author explained, after the detection of the tables a tensor result=results[0] is being generated. This tensor contains all the info for ['bordered', 'borderless'] classes (0,1) which have been classified, as well as their bounded boxes :) If you crop the image containing each table, you can perform OCR to extract the data inside the table.