Vision turns a detection into a motion target
A vision task prepares the scan, runs its detector and goal computer, optionally executes a Cartesian or joint target, and reports a result.
EROBS / DATA FLOW
Vision turns a detection into a motion target
PREPARATION
PERCEPTION
MOTION
RESULT
01
PREPARE
02
DETECT
03
TARGET
04
EXECUTE
05
REPORT
TARGET
Prepare scan
open / move
optional
Detector
image / cached
pose
Goal computer
pose / joints
or return pose
MTC execution
Cartesian /
joints
Action result
pose · success ·
error
Solid: normal data handoff
Dashed: approach_pose with detect_only
Detect Only skips the computed approach. Earlier scan moves can still occur; j6_snap ignores this flag.