How To Use Face Recognition (AI NVR)

How To Use Face Recognition (AI NVR)

Face recognition is used in conjunction with face detection to locate and determine facial similarities detected by the system. Images are registered in a face library, by the user, and accessed via a face recognition smart search tool that is built into the NVR. A USB flash drive with applicable images (in jpeg format) should be inserted into a USB port on the NVR if you are importing images locally.

For more information on how to configure Face Detection, click here.

Step 1: Access the AI menu via the main menu. 

mceclip0.png

Step 2: In the Parameters section, access the Smart Plan menu and click on the Face Detection option. The Face Detection option cannot be used simultaneously with other IVS smart plans. Click Apply (Save).

mceclip0.png

Note: If using a non-AI camera, a smart plan may not be available for the specific channel.

Step 3: Click on the Face Recognition option in the Parameters menu and verify the correct channel (camera) is being used. Click on the Enable toggle switch (checkbox) to enable the feature. Click on Apply (Save).

mceclip0.png

Step 4: Click on the Database option located in the AI menu and click on Add to begin registering images to a face library. A face library must be registered for this feature to function properly. Enter a name for your face library and click Save.

The library will now be saved to the database. Click on the Details section to begin adding images into the to the face library.

mceclip1.png

If you have a single image to add click Register ID, if you have multiple images to add click on Batch register. Follow the on-screen prompts to add the images into the library. Once the images have been successfully modeled into the database left click on the interface to exit.

Step 5: Navigate back to the Face Recognition menu in the Parameters section and click on the Setting option in the Target Face Database field. Select the face library you just created and click Save to insert the library of registered images into the interface.

mceclip2.png

Step 6: Select an AI mode from the AI Mode drop down menu. Common Mode allows the device to use the images registered in the face library as a reference. Enabling the Stranger Alarm will alert the system once an unrecognized face is detected. Click Apply (Save) to save the settings to your system.

mceclip3.png

Note: To adjust Common Mode settings, click on the Trigger (gear) icon and select which additional options you would like to apply to this feature. Be sure to click Apply to save the settings to your device.

How To View Face Recognition Data

All face recognition data retained by the system can be accessed via the Face Recognition option located in the Smart Search menu. A search by attributes or a search by image can be performed. Searching by attributes will load all face recognition data found in the system whereas searching by an image will filter and display only the faces chosen by the user.

To search by attributes, enter a start and end time of the event in the interface and then click Smart Search. The interface will display all face recognition data. To view the event, click on the event you would like to view then click the play button in the interface.

mceclip4.png

To view specific face recognition data by image, click on the Search by Image tab located in the Face Recognition interface.

mceclip5.png

If searchable images are already uploaded into a face library click on Face Library and select an image from the interface, then click Save to begin a search. If you would like to add images locally click Local Upload and follow the on-screen prompts to upload images. A USB flash drive with applicable images (in jpeg format) can be used to upload images to the NVR. Once an image has been loaded, enter a start and end time of the event and click Smart Search.

mceclip6.png

The system will display all face recognition data based on the image selected. All face recognition data can be backed up locally using a USB flash drive.

Was this article helpful?
1 out of 1 found this helpful
Have more questions? Submit a request

Comments

0 comments

Please sign in to leave a comment.