© сайт Томского политехнического университета TOMSK, Apr
9 – RIA Tomsk. Engineer of Laboratory for 3D Modeling and Industrial Design of
Tomsk Polytechnic University (TPU) Stepan Nebaba won grant "UMNIK "
on development of program for face recognition in video stream in real time;
prototype of system is planned to be received in 2019, finish work – in the
2020th, the press service of the university reports on Monday.
It is
specified that the best modern developments in the sphere of face recognition
according to flat images of the road reach high precision of recognition in
difficult conditions of shooting seldom (for example, in crowd). Development of
TPU scientist will allow to combine 2D images in the database with 3D images
received from system sensors including from infrared cameras.
"The
engineer of Laboratory for 3D Modeling and Industrial Design of TPU Stepan
Nebaba became one of winners of the "UMNIK" program of Foundation for
Assistance to Small Innovative Enterprises. The development of the software
product for the multitouch system of recognition of the personality in a video
stream in real time has been awarded", – it is said in the statement.
The system
consists of special sensors for shooting and the software. Sensors are
understood as both video cameras, and cameras of infrared range by which it is
possible to determine distance to an object with rather high precision, making
thus its 3D model.
"My
development is designed to solve the problem of fast non-cooperative face
recognition in the video stream with high accuracy, using the equipment of the
budget price category without involving information processing servers and
strict specifications for sensors," – the press service of the university
cites Nebaba.
According
to the press service, the engineer of TPU plans to receive a pilot prototype of
the system in 2019, to finish work completely
– in 2020. The university doesn't disclose the sum of the received grant
on carrying out development.