Portfolio Category: Video Content Analysis


With our RoboCam project we have intended to develop a device that adds intelligent functions to the capabilities of traditional security cameras using the toolkit of computer vision. Such intelligent functions are, for example, the counting and tracking of persons and vehicles, or the offering of special notification functions. The special feature of RoboCam is that compared to the traditional video analytical solutions it is able to simultaneously handle permanently set and controlled, IP based and analogous cameras.

Visitor Tracking

The person tracking system is monitoring the visitor traffic of shopping centers, institutions and building complexes by processing the images of security cameras. We offer this system to partners who in addition to the traditional visitor information and general security issues need more complex functions: our system supports the analysis of the routes toured by the visitors, the preparation of thermal images, or tracking or retrieving the route of a certain visitor. The user-friendly interface of the system permits data checking any time, even remotely. The system offers real-time and retrievable statistics: even graphical reports can be made according to customized needs.


The CounterGuard passenger, guest, vehicle and visitor counting system is monitoring the incoming and outgoing visitor traffic of supermarkets and clubs from simple catering units with only one entrance up to the complex entertainment locations with several rooms or sites including shopping centers. Our custom-developed new generation guest counting technology gives correct real-time visitor data even for large masses, with the help of stereo camera sensors installed at the access points. The cameras are not recording images, thus the system does not violate the privacy rights. With the user-friendly reporting interface of CounterGuard the client may check the data any time on the computer or remotely on the mobile phone. The system offers real-time and retrievable statistics: customized reports, even graphically, can be compiled for daily, weekly, monthly or specially defined periods.



All over the world more than one hundred million deaf people
are using the sign language, however the real-time
communication between the deaf and hearing people has
remained unsolved until today. In the framework of its long
term development project, Dolphio Technologies is working on
a complex sign language recognition system the realization
of which has been declared impossible by a number of professional
and research workshops. The R+D team of Dolphio Technologies
has been from the beginning convinced that the task is
in-deed manageable. As a result of our hard work, the project
has received international recognition. For example, among
130 projects it qualified for the finals of the Global Impact 2014
competition of the Singularity University CEE founded by NASA
and Google. The SignAll sign language recognition system is
able to make real-time translations from the Hungarian sign
language (HSL) and from one of the most widely used sign language
of the world, the ASL (American Sign Language). The computer
linguistic algorithms-based system is not only able to recognize
the words, but can arrange them into grammatically meaningful
sentences which offers an unprecedented help to people with
hearing disabilities and contributes to the communication
between the deaf-mutes and people who can hear.