Optical Baggage Tracking & Identification

Artificial Intelligence and Computer Vision enable digital tracking of baggage

Computer vision replaces barcodes

Cloudflight developed, validated, deployed and hosted the software systems for optical baggage tracking and identification, leveraging the vast variety of in-house experts. We adapted state of the art machine learning approaches from other identification problem domains, e.g. face recognition, to the requirements of baggage identification.

Alessio Montuoro, Senior Data Scientist at Cloudflight

Alessio Montuoro
Computer Vision


faster onboarding for flight passengers

Baggage tracking at the airport: from drop-off via security check until baggage reclaim, every suitcase is tracked and can be identified.


Tracking baggage using conventional means (like bar codes etc.) is both costly and error prone, for example, bar codes can be lost. Furthermore, a continuous tracking of baggage within large scale installations requires a large number of tracking sensors – which are both costly in acquisition and maintenance.


By augmenting existing tracking solution and camera based systems, the overall tracking accuracy is improved and new use cases can be covered.

Camera systems mounted at key locations of the baggage transport logistic process (e.g. at check in counters, before and after security screenings, at aircraft loading bays, …) can both detect and track baggage and thus ensure a complete audit trail.

This can be applied in a variety of use cases:

  • Security: Was the baggage being loaded into the aircraft handled at the security screening checkpoint?
  • Lost and found: Did the checked in baggage make it into the aircraft?
  • Liability: Where was a baggage lost and by whom? At the airport? By the airline?
  • Theft prevention: Did a baggage item unloaded from a plane arrive at the baggage reclaim area? If so, how long was it present there?


Instead of developing a large centralized system, Cloudflight divided the solution into two main parts:

  • A small and cost effective camera and image processing services
  • and a centralized identification and tracking service

Using this architecture we are processing images in near real time. This means that the video doesn’t need to be streamed over the network, what contributes to low network load. Also, central processing server is not needed (no single point of failure).

Image Processing Service

Cloudflight created a multi stage real time image processing pipeline consisting of the following steps:

Object detection: detection, localization and classification of baggage on the video stream

  • What objects are visible?
  • Where are these objects located on the video?
  • What type of objects are visible? Persons? Baggage? Cargo container?

Instance segmentation: which parts of the image belong to which object exactly?

  • Multiple, potentially touching and overlapping baggage?
  • Persons handling and occluding baggage items?

Object Fingerprinting: finding a numerical and representation of the characteristic of the detected baggage

  • What features distinguish the baggage most from others?
  • There are millions of black trolleys, but maybe only one has a red Catalysts logo on it?

Identification and Tracking Service

Central server receives object fingerprints from image processing services and stores them in a central database. By comparing these fingerprints – a quick and easy operation – the server can easily compare current detection of baggage items to previous detection and thus build a complete audit trail.

Automated Baggage Scanning


  • Do you need to track individual items in your production plant?
  • Would your business benefit from real time tracking of assets?
  • Automated alerts when specific objects leave or enter virtual geo fences?

Alessio Montuoro
Computer Vision

Sign up to our monthly wrap-up and get access to exclusive material and updates about our technologies, research content and events. You can always adjust your settings.

Sign-up form by MailChimp