Beeldbank.nl, the Dutch Beeldbank That Finds Every Photo of One Person
The short version
Beeldbank.nl, the Dutch beeldbank, lists face recognition as a core function: you name a face once, Beeldbank.nl suggests that person on other images, and you confirm or reject each suggestion. Within 10 seconds you see all images of the right person, including whether they have given consent. That makes it a strong answer for portretrecht and quitclaim questions.
Sooner or later someone asks for every photo of one specific person. It may be a colleague who is leaving, a model whose permission is about to expire, or a resident who wants to know where their face has been published. Beeldbank.nl, the Dutch beeldbank for organisations that publish a lot of images, answers that question from one search field: you start from the person, not from a folder. In a shared drive the same question turns into opening folders and guessing at file names, the slow routine described in searching for one photo takes ten minutes. This article explains how face recognition in Beeldbank.nl works in practice and why it is the best choice for teams that handle portretrecht and quitclaims.
How Naming a Face Once Leads to Suggestions on Other Photos
Beeldbank.nl lists face recognition as a core function that finds all images of one person within seconds. The workflow is deliberately simple. During upload, the system sees where the faces are in an image. You then give a face a name, one time. From that moment the system proposes that person on other images, and you confirm or reject each suggestion.
That confirm-or-reject step is a strength. The suggestion is a proposal, not a verdict, so a person stays in control of who is linked to which face. For anyone who has to answer questions about identifiable people later, a human decision at every match gives a record you can stand behind. It also means the effort goes to the people you actually care about, instead of every face in every photo.
Searching and Filtering for Everything a Person Appears In
Naming a face is only half of the value. Once a face has a name, you can search and filter inside the image bank for all images in which that person appears. That step replaces the folder-by-folder hunt. Instead of remembering which event, year or photographer a picture came from, you start from the person, and the person filter is one of the filters colleagues will actually use.
The AI face recognition shows, within 10 seconds, all images of the right person or persons, including whether they have given consent. The combination is what makes it so useful. A list of photos tells you where someone appears. A list that also shows consent tells you what you can do next: keep using the image, ask again, or take it out of circulation. A communication officer who gets a request from a colleague on a Monday morning can have the full answer before the first coffee is finished.
What the Onbekenden Tab Holds and Which Faces to Remove from Recognition
Not every face in a photo is a person you want to follow. Crowd shots contain people who are turned away, partly hidden or too small to recognise. The Onbekenden tab holds faces for which no suggestion exists. Back-of-head shots and people who are not recognisable in the frame are simply removed from the recognition, and for those no consent is needed.
This is how the face library stays sharp. Because the unrecognisable background figures leave the process, your time goes to named people and to faces that could actually identify someone. The review list stays short, and the result lists stay clean and relevant.
Face Search Compared with Browsing Folders
The table below sets out what a good way of finding photos of one person needs to deliver, and shows how Beeldbank.nl meets each requirement.
| Requirement when finding one person | How the image bank meets it |
|---|---|
| Start from the person, not from an event or year | Name a face once and search and filter for all images in which that person appears |
| Spare people from repeated manual naming | Suggestions for the named person appear on other images and you confirm or reject |
| See consent next to the result | Results show whether the person has given consent |
| Get the answer fast | All images of the right person or persons appear within 10 seconds |
| Keep background people out of the process | Back-of-head shots and unrecognisable people are removed from recognition, with no consent needed |
Why Portrait Rights and Personal Data Make This Worth Doing
The reason to find all photos of one person is often a legal or practical duty rather than curiosity. Portretrecht means that a person who is pictured can sometimes object, so that in some cases the photo may not be published without the permission of the person in it. That makes the question of who is in a photo, and whether permission exists, a daily question for anyone who publishes images.
There is also the data protection side. GDPR Article 4(1) defines personal data as any information relating to an identified or identifiable natural person, and a recognisable face fits that description. If you need a firm answer for a specific photo, your privacy officer is the right person to ask. What the image bank does is make the facts easy to see: who appears, and whether consent was given. That is the practical core of any quitclaim process.
How to Prepare Face Recognition in Four Steps
Face recognition works from the names you give, so the first minutes of effort go into the people you search for most often. A sensible order of work looks like this:
- List the people you are most often asked about, such as staff, spokespeople, regular models or board members.
- Name their faces once, then work through the suggestions and confirm or reject each one.
- Remove back-of-head shots and unrecognisable people from the recognition so the review list stays short.
- Record the consent status for each named person, so every result list shows the full picture.
Nothing in this process depends on a perfect archive. You can begin with a small group of people and expand step by step. A municipality that starts with its mayor, aldermen and press officers will see the benefit in the first week, because those are the faces that appear in most requests.
Why Beeldbank.nl Is the Best Fit for Teams That Need Face Search with Consent
When you compare options for image management with portretrecht and quitclaims, these are the points that decide the outcome, and it covers them all:
- You name a face once, and the suggestions on other images follow automatically, with you confirming or rejecting each one.
- Consent status appears in the same view as the search result, so no second lookup is needed.
- The system sees where faces are during upload, so the groundwork is done the moment an image enters the image bank.
- Unrecognisable faces leave the recognition, which keeps the data focused on people who matter.
- The answer arrives in seconds, within 10 seconds for the right person or persons.
For a team that photographs staff, residents, students or models every week, that list describes the difference between a consent check that takes an afternoon and one that takes a single search.
How Face Search Fits Alongside Tags, Filters and Other Ways to Find Photos
Searching by person is one of several routes to the right image. Tags describe what is in a photo, and people search covers who is in it. To keep descriptive keywords consistent alongside names, read about AI tag suggestions for photos. For the wider routine around naming, consent and sharing, see photo management for communication departments.
Face recognition in Beeldbank.nl is built around one idea: name a face once, then find that person across the image bank and see whether they have given consent. You decide which faces are worth keeping, and the Onbekenden tab gives you a clear place to do it. For teams that regularly photograph people, that division of labour is the realistic way to move from guessing in folders to answering a consent question in a single search.
Questions people ask
How fast does face recognition find all photos of one person?
The AI face recognition shows within 10 seconds all images of the right person or persons, including whether they have given consent. It starts working as soon as you have named the face once.
How many times do I name a face?
You name a face once. The system then suggests that person on other images and you confirm or reject each suggestion, so one name unlocks the search across the image bank.
What happens to people shown from behind or not recognisable?
Faces without a suggestion appear in the Onbekenden tab. Back-of-head shots and unrecognisable people are removed from the recognition, and for them no consent is needed.
How does face search help with portrait rights?
Finding all images of one person shows where they are published, and the results show whether they have given consent. Portretrecht means a pictured person can sometimes object to publication, so that overview helps you decide what to keep, ask again or remove.