Video recognition algorithms for the automotive industry

Through the last years LANIT-TERCOM has experienced significant improvement in video recognition technologies development. This progress would be impossible without mathematical software and algorithms development.

One of the solutions by LANIT-TERCOM has the following features:

Video capture subsystem, connected with the two high-resolution cameras
Predicts dangerous situations (e.g. collisions or other emergencies) by the motion dynamics analysis
Initial video processing subsystem – level correction and normalization, image rectiﬁcation for the right and left cameras
Image segmentation subsystem
Global parameters search subsystem (e.g. for automotive market – skyline, linear and angular acceleration and camera speed)
The subsystem of the sample objects with the existing model – road signs, cars, traﬃc lane markings