I've been exploring your project, and I had a question about the bounding box (bbox) generation for each element detected. I'm wondering if the bbox is intended to encompass the entire page or if any post-processing steps are applied.
I've been contemplating enhancing the layout detection to integrate it with another tool, as occasionally, elements like lists are recognized as paragraphs. My idea was to align the bounding boxes accordingly, but I've noticed that they don't align consistently, which seems unusual. I would appreciate your insight on this matter.
Is there any way of obtaining the real position relative to the page @ansukla @ianschmitz ?
Hi,
I've been exploring your project, and I had a question about the bounding box (bbox) generation for each element detected. I'm wondering if the bbox is intended to encompass the entire page or if any post-processing steps are applied.
I've been contemplating enhancing the layout detection to integrate it with another tool, as occasionally, elements like lists are recognized as paragraphs. My idea was to align the bounding boxes accordingly, but I've noticed that they don't align consistently, which seems unusual. I would appreciate your insight on this matter.
Is there any way of obtaining the real position relative to the page @ansukla @ianschmitz ?