# Multi-Camera Person Tracking: How to Maintain a Persistent ID Across Non-Overlapping CCTV Cameras?

**URL:** https://forum.opencv.org/t/multi-camera-person-tracking-how-to-maintain-a-persistent-id-across-non-overlapping-cctv-cameras/24718
**Category:** Python
**Created:** [September 21, 2026, 7:47am UTC](https://forum.opencv.org/t/multi-camera-person-tracking-how-to-maintain-a-persistent-id-across-non-overlapping-cctv-cameras/24718 "2026-09-21T07:47:26Z")
**Posts on this page:** 1
**Page:** 1

<div class="post-metadata">

### Author: ![Samir\_Durga](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.opencv.org/samir_durga/32/14481_2.png) [@Samir\_Durga](https://forum.opencv.org/u/Samir_Durga)
#### Post date: [September 21, 2026, 7:47am UTC](https://forum.opencv.org/t/multi-camera-person-tracking-how-to-maintain-a-persistent-id-across-non-overlapping-cctv-cameras/24718/1 "2026-09-21T07:47:26Z")

</div>

Hello everyone,

I am working on a retail/shop surveillance project where I need to track customers across multiple CCTV cameras.

Currently, each camera has its own live video stream, and I am using **YOLOv8 for person detection and tracking with bytetrack**. Within an individual camera, the tracker assigns a local tracking ID to each detected person.

The problem is that these IDs are camera-specific. When the same customer leaves Camera 1 and later appears in Camera 2, the tracker assigns a completely new ID.

My actual requirement

I want to maintain a **single persistent ID for the same person across all cameras** , even though the cameras have **non-overlapping fields of view**.