Videntifier is a visual search engine based on a patented large-scale local feature database. Around it, we have built powerful digital forensics services for law enforcement and others. We do fast, robust identification at low cost in vast visual feature databases.





Lightning fast. Incredibly robust.

Videntifier creates unique visual descriptors from images and videos. These visual interest points are compared with a database to find visually similar matches. The process is extremely fast, reliable, and cost-effective.

Query Video
Visual interest points are identified

Video in Database
Interest points match to the database


With the powerful Videntifier database technology, we are able to use highly fine-grained and robust interest point identification. Therefore, Videntifier easily recognizes modified content, finds partial resemblances, and has amazingly low false positive rates.

For a brief technical whitepaper on the technology please contact us below


EXTREME performance without COMPROMISE


high number of descriptors enables the technology to tolerate significant modifications to the content, meaning that it is very difficult to fool ii

Extreme scalability

databases that contain up to billions of images, millions of hours of videos and thousands of live streams without noticeable change in search time


ability to search through hundreds of millions of images per day, with very little effect on performanc

Privacy by design 

all content is one way encoded and it cannot be recreated using the descriptors


The solution’s structure allows it to run on cost efficient hardware without loosing performance 

Partners and funding sources

This project has RECEIVED funding from the Eurostars-2 joint programme with co-funding from the european union Horizon 2020 research and innovation programme and was promoted as a Eureka success story by the eureka secretariat .



Does videntifier recognize faces?

Videntifier is not optimized to recognize faces. However the technology has been used to identify passport photo duplicates in very large image collections.

Does Videntifier recognize objects?

The base videntifier search engine has limited semantic recollection. Videntifier will easily match together most logos and buildings as well as many distinct objects. Videntifier is currently developing a semantic search engine in cooperation with partners.

Does Videntifier save the content it processes?

Videntifier creates and processes a unique visual descriptor for each file which cannot be turned back into content. This guarantees that Videntifier is compliant with the strictest privacy demands.

What types of modifications does Videntifier handle?

Due to the high number of descriptors used by the Videntifier search engine it is incredibly robust against most forms of modifications. Some of the modifications include, cropping, flipping, rotating, blurring, color changes, partial overlays and more.

How much content can Videntifier identify per day?

The extremely high performance database structure used by Videntifier is able to visually identify an incredibly high amount of content without compromising on accuracy or robustness. Solutions that identify billions of images per day or hundreds of live streams are available as a cost effective service.

How is my content kept secure?

Videntifier applies one way encoded visual descriptors that cannot be turned back into content.