Applies to Product:
Applies to Firmware: 8.2.2 and newer
Face detection and recognition is an advanced AI feature on X-Series devices that allows users to define a custom perimeter within the camera’s field of view (FOV) to detect and identify faces. This guide provides a streamlined, professional overview of configuring this feature, building on the face database management setup to ensure accurate detection, recognition, and event triggering.
Step 1: Access the Device
Log in to your X-Series device using your administrator credentials.
Step 2: Navigate to Face Detection Settings
From the Main Menu, select Setup > AI > Setup > Face Detection tab.
Step 3: Configure Face Detection Settings
Adjust the following settings to meet your requirements:
Read & Agree: Check the box to enable the face detection feature.
Channel: Select the camera channel for face detection configuration.
Enable: Toggle to activate face detection for the selected channel.
Face Features: Enable to detect attributes such as masks or glasses within the field of view (FOV).
Snap Mode: Choose how snapshots of recognized faces are captured:
Realtime Mode: Captures faces entering and leaving the detection area, sending two notifications (entry and exit) to the Alarm Notification Panel for continuous monitoring.
Optimal Mode: Captures a single, high-quality snapshot of the face.
Interval Mode: Specify the number of snapshots and the time interval between them.
Apply Mode: Select the facial recognition scanning method:
Frontal View: Optimizes scanning for faces approaching the camera head-on.
Multi Angle: Optimizes scanning for faces from various angles.
Customize: Allows custom angle configurations for face scanning.
Min Pixel & Max Pixel: Set the minimum and maximum pixel dimensions for detected faces to refine crowd density monitoring. Higher values detect larger faces; lower values include smaller faces.
Detection Mode: Configure detection for pedestrians, vehicles, or non-motorized vehicles:
Motion Mode: Detects only moving objects, ideal for high-traffic areas.
Hybrid Mode: Detects both moving and stationary objects, suitable for comprehensive monitoring (e.g., parking lots).
Trigger Mode: Define the detection area:
Rectangle: Draw a rectangular region in the live video feed to restrict face detection to that area.
Line: Define a virtual line in the feed, with options to specify movement direction (e.g., left to right, right to left, or both) for triggering face detection.
Detection Area (Rectangle Trigger Mode only):
User-Defined: Draw a custom area for crowd density monitoring.
Full Screen: Monitor the entire camera FOV.
Draw Rule: In the live video feed, click to set points and create a perimeter shape. Verify the rule type and direction align with your requirements.
Step 4: Save Settings
Click Save to apply the configured settings.
_____________________________________________________________________________________
Step 5: Configure Face Detection Schedule
To schedule when Face Detection is active:
Navigate to Main Menu > Setup > AI > Setup > Schedule.
Select the Channel for scheduling.
Enable scheduling for the selected channel.
Choose Face Detection from the list of AI features.
In the schedule interface, click and drag to select days and hours for Face Detection activation.
(Optional) Use the Copy option to apply the schedule to other channels or configure each channel individually.
Click Save to confirm.
_____________________________________________________________________________________
Step 6: Configure Face Database Management
The face database is essential for managing flagged, non-flagged, and unrecognized faces.
Navigate to Main Menu > Setup > AI > Recognition > Face Database Management.
Default groups include:
Allow List: Add faces of authorized individuals (e.g., employees) who do not require alerts.
Block List: Add faces of individuals to flag (e.g., persons of interest, such as those banned for shoplifting).
Stranger List: Track unrecognized faces to trigger alerts for unauthorized individuals.
Enable the desired group lists.
4. Click the Edit icon next to each group to configure its database.
5. Select Import in the bottom corner of the edit page.
6. Choose an import method:
Capture Image: Import faces detected by the X-Series device and create profiles in the corresponding database.
Local Image: Upload images from a USB flash drive or computer to the database.
7. Click Save to complete the configuration.
_____________________________________________________________________________________
Step 7: Configure Triggered Actions (Optional)
Set up actions to trigger when a face is detected or recognized:
Navigate to Main Menu > AI > Alarm > Face Recognition.
For each group, configure the following options:
Enable: Activate triggered actions for the group.
Similarity: Adjust the percentage threshold for a face to be considered a match (lower values allow less precise matches; higher values require closer matches).
Alarm: Configure actions to trigger upon face detection:
Channel: Select the camera channel for alarm settings.
Buzzer: Enable a beep from the XVR/NVR upon detection.
Latch Time: Set the duration (in seconds) for the buzzer to sound.
Face Capture: Enable to save snapshots of recognized faces to the database for improved recognition accuracy.
Save Background: Enable to save the background with the face snapshot.
Show Thumbnail: Enable to display a snapshot in the Alarm Notification Panel in Live View mode.
Send Email: Enable email notifications (requires email configuration in Network Settings).
FTP Picture Upload: Enable image uploads to an FTP server (requires FTP configuration in Storage Settings).
Cloud Picture: Enable image uploads to a cloud storage service (requires cloud configuration in Storage Settings).
Alarm Schedule: Set days and hours for alarms to be active (default is 24/7; adjust as needed).
Alarm Channel: Select channels to trigger alerts when faces are detected.
Click Save to confirm the alarm settings.
Related to
Comments
0 comments
Please sign in to leave a comment.