Skip to main content
RecogResultData is typically returned as the recognizeResult field of Element. It describes the recognized category and key point location information.

Purpose

You can treat this as a “recognition output summary”, which typically includes:
  • Recognized category: predict_name indicates which category the model/rules think the element belongs to (default: others).
  • Recognition region: up_left_point_* and down_right_point_* form a bounding box that represents the region on the page.
  • Key point: key_point_* is a representative anchor point used for later interactions or positioning.
These fields use pixel coordinates, which makes them convenient for direct UI display and interaction positioning.
FieldTypeDescription
predict_namestringCategory name (default: others)
up_left_point_xnumberTop-left X (pixel coordinates)
up_left_point_ynumberTop-left Y (pixel coordinates)
key_point_xnumberKey point X (pixel coordinates)
key_point_ynumberKey point Y (pixel coordinates)
down_right_point_xnumberBottom-right X (pixel coordinates)
down_right_point_ynumberBottom-right Y (pixel coordinates)