New posts in Hocr

How do I segment a document using Tesseract then output the resulting bounding boxes and labels

Does Tesseract's hOCR output really contain bounding boxes and confidence levels for each character?