Participant identification in endurance-event galleries

Runners at a large race often hope to open a gallery after the event and find themselves in a few clicks. That is a reasonable hope. Making it true depends on how images are matched to people, and on what is reasonable to promise.

This page belongs to our guide to running, cycling and endurance-event photography.

How can photographs be matched to participants?

Several methods exist, and events often combine them.

• Bib numbers. A visible number is read from the image, either by a person or by software, and linked to a participant.
• Timing records. Finish times or checkpoint times narrow the search to a time window.
• Registration lists. A list of names, numbers and categories supports sorting by wave or club.
• Manual sorting. A person looks at the images and matches them, which is accurate but slow.

None of these is perfect on its own. Each has strengths, and each fails in some conditions.

What limits accuracy?

Bibs are covered by arms, jackets or hydration vests. Numbers are folded, wet or not worn on the front. Crowded finish areas hide runners behind others. A photograph taken from the side may not show the number at all.

So even a good system leaves some images unmatched. A promise that every participant will be found is not one we can responsibly make. We describe the likely rate of success for your course and conditions, and we say that some images will need manual review.

What are the privacy questions?

A gallery that links names to photographs involves personal information. Malaysia's Personal Data Protection Act 2010 regulates the handling of personal data in commercial transactions, according to the Personal Data Protection Commissioner's own page. Whether and how it applies to your event is a legal question, and we are not in a position to give legal advice.

We suggest you take qualified advice, and that you agree in writing who holds participant lists, who may see them and how long they are kept. Contracts, venue rules and your own policies may add further requirements. Our guide to privacy, permission and image use covers the practical questions.

What about automated facial matching?

Some systems claim to find a participant from a face. Their reliability, the data they require and the privacy questions they raise vary. We do not assume that any such tool is suitable or available for your event.

Read facial recognition and automated guest-photo search before considering one. If you are interested, ask us whether this can be included; we confirm what is actually possible in writing, after checking the system's own documentation.

A short example

For example, imagine a 10-kilometre run with 3,000 participants, bibs pinned to the chest and a timing provider that records finish times. The organiser wants participants to search by bib number.

A plan might sort images by finish time first, then read bibs where visible, with a manual review of unmatched images. The organiser accepts that a small share may remain unmatched and sets up a way to report a missing photograph. This is a hypothetical plan, not a past job.

Questions to ask any gallery provider

• How does the system read or match numbers, and what share of images does it usually leave unmatched?
• Who can see the participant list, and where is it stored?
• Can a runner ask for an image to be hidden or removed, and how quickly?
• What does the provider's own documentation say about its limits?

Ask for answers in writing. Claims about a system are only as good as its published documentation and a trial on your own course.

What we confirm in writing

Your quotation states the sorting method, the data involved, who holds it and how mismatches are handled. For the related case of hidden numbers, read race photographs when bib numbers are hidden or missing. For working with a results company, see photography coordinated with a race timing provider. The photo sharing and privacy plan helps you record the decisions.