The result of one OCR run. detections is always the flat list of
recognized regions; blocks and regions carry the layout grouping (one
'ungrouped' block without a layout model); image is the frame all
coordinates are relative to: the corrected page after any orientation or
dewarp pass, or the input.
The result of one OCR run.
detectionsis always the flat list of recognized regions;blocksandregionscarry the layout grouping (one'ungrouped'block without a layout model);imageis the frame all coordinates are relative to: the corrected page after any orientation or dewarp pass, or the input.