Skip to main content
Version: V12

How to Manage the Object Library

The Object Library is where you enroll the people, vehicles, and objects you want your cameras to recognize. This guide covers adding subjects, improving them with more photos, editing their details, finding them, reviewing where they have been seen, and removing them.

Prerequisites

  • Ensure you belong to a group with the Object Library feature enabled, or have a CAL that grants its permissions. Adding, editing, and deleting subjects are each gated separately.
  • Have your reference photos ready as JPEG, PNG, or WebP images. Use clear, well-lit images where the subject is the obvious focus.

Note: Recognition only works when the matching object type is being detected on your cameras. See How Live AI Detection Works to set up detection.

Enroll a New Subject

Step 1: Open the Object Library

  1. Navigate to the Object Library in the surveillance area of your portal.
  2. Select Add to Library. (When the library is empty, select Add First Subject.)

Step 2: Choose the Type and Enter Details

  1. Select a Type: Face, Person, Vehicle, or Object.
  2. Enter a Name for the subject. This is required.
  3. Optionally enter an External ID to link the subject to a record in another system.
  4. Optionally enter a Description.
  5. Set a Watchlist Priority: High, Medium, or Low.
  6. Optionally add Tags to help you organize and search the gallery.

Important: The Type is fixed once the subject is enrolled and cannot be changed later. To correct it, delete the subject and enroll it again.

Note: Live recognition currently matches Face subjects against camera feeds. You can enroll the other types to organize the library; recognition for Person, Vehicle, and Object is planned and will use the same enrollment flow. See Understanding Face Recognition.

Step 3: Add Reference Photos

  1. Select Select Photos and choose one or more images.
  2. Review each preview. The system detects the subject, outlines it, and shows a Detection Score and Quality Score badge for each photo.
  3. If a photo is flagged as low quality but you still want to use it, select Accept marginal quality.
  4. Select Enroll Subject.

The subject now appears in the library. If any photo failed, it stays in the dialog so you can adjust and retry without re-entering the rest.

Note: More good photos from different angles and lighting generally improve recognition accuracy.

Add More Photos to an Existing Subject

  1. Select the subject to open its detail panel.
  2. Under Reference Photos, select the upload area (Add photos).
  3. Choose the additional images and select Add Photos.

Edit a Subject's Details

  1. Select the subject to open its detail panel.
  2. Select Edit Details (the pencil icon).
  3. Update the Name, External ID, Description, Watchlist Priority, or Tags.
  4. Select Save Changes.

Note: Editing changes details only. It does not re-process the photos, and the Type stays fixed.

Find Subjects

  • Use the category filter pills at the top (All, Face, Person, Vehicle, Object) to narrow the gallery by type. Each pill shows a count.
  • Use the search box in the top bar to find subjects by name, description, external ID, or tag.
  • Use the view control to switch between thumbnail and list layouts.
  • Select Refresh to reload the gallery.

Review Recognition Activity

  1. Select a subject to open its detail panel.
  2. Under Recognition Activity, review where and when the subject was matched. Each entry shows the recording, the time, and the match confidence.
  3. Select an entry to open the recording at that moment.

A subject's card also shows a Last Seen line when it has recent activity, or indicates it has not been seen yet.

Remove Photos or Subjects

To remove a single reference photo:

  1. Open the subject's detail panel.
  2. Under Reference Photos, select the delete icon on the photo.

To remove an entire subject:

  1. Open the subject's detail panel.
  2. Select Delete Subject.

Note: Deleting the last remaining photo of a subject retires the subject. A removed subject stops being recognized going forward.

See Also