Set Up Line-Crossing Detection Event
Setting up a line-crossing detection event detects people who cross configured virtual lines and triggers real-time events. Specify the crossing direction (A→B, B→A, or both) to monitor a specific direction only. Optionally add a face area to prioritize face tracking; enrolled users are automatically identified.
When to use this?
-
Monitor and count doorway traffic
-
Monitor entry in a specific direction only (detect one-way traffic)
-
Detect entry into specific areas outside business hours
-
Track left/right area movement relative to the line
Create line-crossing detection event
Use the Live menu to detect people crossing a configured line in a specific area and trigger events. Also, setting a face area prioritizes face tracking in that area; enrolled users are automatically identified.
Start event creation
-
Click the AI&Rule menu in the upper-left corner of the screen.
-
Click AI Analysis on the left sidebar.
-
Click + Add at the top right of the screen.
-
In the Select AI Analysis Type window, click Line Crossing.
-
When the Notice and Consent for Activation of Vionyx Facial Recognition Function pop-up appears, review the details, select the consent checkbox, and click Agree and Activate Facial Recognition.
-
To use the Facial Authentication feature, comply with local regulations. Cannot proceed without consent.
-
The Notice and Consent for Activation of Vionyx Facial Recognition Function popup appears only once. Subsequent AI event creation assumes consent.
-
-
Complete the required settings in the right Line Crossing panel.
Set event name
Enter the AI event name in the Name input field. The name can be up to 64 characters long.
Virtual line settings
Place the virtual line at the appropriate location in Live View. The camera detects people crossing configured virtual lines and triggers events. Set at least one virtual line. Click + Add to add up to 10. An orange line appears in Live View; drag corners or points to adjust the shape. Drag the center to move the line.
-
Name: Edit the virtual line's name. Assign a clear, descriptive name. The virtual line name appears on hover in Live View.
-
Condition: Set the crossing condition for the virtual line. Select one of the three direction options below.
-
A -> B: Trigger an event when crossing from A to B.
-
B -> A: Trigger an event when crossing from B to A.
-
A <-> B: Trigger an event when crossing in both directions.
-
-
Reference Point: Select the reference point in the person detection box to determine whether a person crosses a virtual line. Set this per line. The default is Center.
-
Center: Use the center of the box as the reference point. The system counts a crossing when the body center touches the line.
-
Lower: Use the point 80% down from the top along the box’s long axis. The system counts a crossing when the lower body touches the line. Select this option to trigger an event when feet cross the line, as if it were on the floor.
Click the icon to the right of Reference Point to view the location and detection criteria for each reference point in a diagram.
-
-
Support for the Reference Point option is included in firmware version 1.1.0 or later. On devices that do not support reference point selection, all line crossings use Center.
-
Lower is a position calculated from the person detection box, not the actual foot position.
-
Enable Reference Point in AI display options to check the reference point in the Live menu.
-
Click to reset the configured virtual line's shape and size.
-
Click to delete an added virtual line.
Face area settings (optional)
Define the area for face detection. When set, the system prioritizes face tracking in that area. Face areas can overlap with virtual lines. Click + Add to add up to 10 face areas. A blue box appears in Live View; drag corners or points to adjust the shape. Drag the center to move the line.
- Name: Edit the face area's name. Assign a clear, descriptive name. The face area name appears on hover in Live View.
-
Enrolled users are automatically identified in the Live menu and shown by name. Unenrolled users appear as Unknown Face.
-
Click to reset the configured face area's shape and size.
-
Click to delete an added face area.
-
In camera settings, if you use the Adaptive AE mode option, configure it so that the face area includes the position where a person would actually stand or pass by and their face is clearly visible.
Person detection settings (optional)
In the Detection section, select the area where the camera detects people. Narrow the detection area as needed to reduce false detections in irrelevant areas (hallway ends, windows, reflective surfaces, etc.).
-
All Area: Default. Detects people across the entire screen. Available regardless of whether face areas are set.
-
Designated Area: Detects people only within the area you specify. Available only when no face area is set. Set at least one virtual line. Click + Add to add up to 10. A purple box appears in the live view; drag its corners or points with the mouse to adjust it to the desired shape. Drag the center to move the line.
- Name: Edit the name of the person detection area.
-
Face Area Only: Detects people only within the face area. Available only when a face area is set.
-
Adding a face area while Designated Area is selected opens the Notice pop-up. Click OK to clear the designated detection area and switch to Face Area Only.
-
Deleting all face areas while Face Area Only is selected automatically switches to All Area.
-
In camera settings, when you use the Adaptive AE mode option, narrow the range to Designated Area or Face Area Only. Keep screens, photos, or posters with faces out of the detection area.
-
Click to initialize the shape and size of the person detection area.
-
Click to delete the added person detection area.
-
Firmware version 1.1.0 or later supports this feature.
Set event trigger conditions and event count settings
In the Option section, set the event trigger conditions and event count.
-
Every Delay: Set the minimum duration to treat as an event. An event occurs only when the detection state lasts longer than the set time. It can be set up to a maximum of 60 seconds.
If too many events occur, increase the Every Delay value. Motion detected for less than the set time is not treated as an event, reducing unnecessary events.
-
Event Count: Set to On to count each line crossing. If multiple lines exist, count each line separately.
-
All Event Count Reset: Set to On to initialize event counts for all lines daily at the specified time. Initialize event counts for all lines at the time set in Reset Time.
-
If you set Event Count to On, it is displayed in the Live menu above the video for each line in the Count : N format. Click the button and select Count Reset to initialize the event count manually.
-
Support for event count options is included in firmware version 1.1.0 or later.
Save settings
After completing all settings, click Create in the top right. The interface returns to the previous screen, and the new AI event is added to the list.
Check in Live
Click the Live menu at the top left of the screen. Confirm the 
icon at the top right of the camera feed is on. Configured virtual lines and face areas are overlaid on the video. An event occurs each time a person crosses a virtual line. If face areas are set, the camera detects faces there; enrolled users are automatically identified and their names are displayed.
-
Unenrolled users appear as Unknown Face. Enrolled users appear with their name.
-
Poor lighting or backlight may prevent the camera from detecting faces. Improve area lighting or adjust the camera position and angle so faces are clearly visible.
-
Faces must face the camera to be detected. Profile or rear views may not be detected.