Download IVISION Catalog
01Prepare the view
Define the permitted cameras and review interval.
Evaluate person re-identification across views
Camera angle, clothing changes and lookalikes can undermine associations. Verify the specific method and use source images to confirm or reject proposed links; a re-identification score does not establish a person’s name or continuous route.
Catalog workflow. Confirm detector availability, permitted use and event delivery for the proposed deployment. Website examples are illustrative.
Assess re-identification methods
Example
Without IVISIONStandard CCTV feed
With IVISIONIllustrative event overlayThe Process
How It Works
Similar appearance is a candidate link Make cross-view associations accountable.
01
02Assess the observation
Validate matching under actual angle and lighting changes.
03Review the evidence
Compare each proposed pair with source context.
04Choose a response
Retain only supported links and document uncertainty.
Explore Your Options
Explore Related Capabilities
| Dimension |  Person re-identification across different cameras and locations |  multi camera face tracking |
|---|---|---|
| Observation |  Sightings with visually similar supported characteristics. | Possible sightings of the same enrolled face. |
| Event criteria | A validated association method proposes a cross-view match. | Validated matching proposes a link between observations. |
| Monitoring setup | Confirm camera scope, time window, features and threshold. | Confirm camera scope, time window and matching thresholds. |
| What it cannot establish | Cannot establish legal identity, intent or unseen movement. | Blind spots and false matches can break an inferred route. |
| Practical outcome | Help reviewers compare sightings without treating appearance similarity as identity. | Compare separate sightings without implying continuous visibility. |
Illustrative demonstration
Review The Event.
Understand The Context.
Example footage and simulated events explain the review workflow. They do not measure detector performance.
Sites overview
Example Activity
No matching events.
Related Scenarios
View MoreTracking & Movementmulti camera face tracking
Evaluate a face-specific association approach.
Read More
Tracking & MovementCross-camera route reconstruction
Use confirmed links in a route review.
Read More
Tracking & MovementCreation of visit histories for unknown or unidentified individuals
Assess governance before storing repeated unidentified links.
Read MoreEXPLORE WHAT’S POSSIBLE
What Can You Do With Person re-identification across different cameras and locations?
01 / 04 · PERSON RE-IDENTIFICATION ACROSS DIFFERENT CAMERAS AND LOCATIONS
Logistics
Uniforms create many lookalikes.
Example event: A cross-view candidate appears.
Operator response: Check timing and distinct visible evidence.
Practical value: Avoids clothing-only attribution.
Discuss this capability
02 / 04 · PERSON RE-IDENTIFICATION ACROSS DIFFERENT CAMERAS AND LOCATIONS
Offices
A person changes outerwear.
Example event: The association confidence changes.
Operator response: Review alternatives rather than forcing a match.
Practical value: Preserves uncertainty.
Discuss this capability
03 / 04 · PERSON RE-IDENTIFICATION ACROSS DIFFERENT CAMERAS AND LOCATIONS
Industrial security
A transition includes a blind interval.
Example event: Two sightings are linked provisionally.
Operator response: Confirm feasible timing and source images.
Practical value: Supports cautious reconstruction.
Discuss this capability
04 / 04 · PERSON RE-IDENTIFICATION ACROSS DIFFERENT CAMERAS AND LOCATIONS
Deployment review
Use views with useful body or face detail for the approved method.
Visibility: Occlusion and appearance changes degrade links.
Configuration: No specific re-identification engine was verified.
Privacy: Restrict persistent cross-camera histories to justified uses.
Human review: Reject uncertain links rather than filling gaps.
Discuss this capability
Frequently asked questions
Person re-identification across different cameras and locations, Explained.
Associates candidate person sightings across views using a validated visual matching method.
Is re-identification the same as face recognition?
Not necessarily. The deployed method may use different visual evidence.
Does it reveal someone’s name?
No. Identity needs an independently verified association.
Can uniforms cause confusion?
Yes. Similar appearance is a key validation case.
Can a score prove a match?
No. Review the evidence and method limits.
Find your fit
What are you trying
to achieve?
Help reviewers compare sightings without treating appearance similarity as identity. For Authorized investigators, Video analytics evaluators.
Workflow to discuss
Person re-identification across different cameras and locations + Review associations individually.
