Computer Vision Systems
The company brings in world-class experts to work on innovative machine vision solutions built on the stereo principle and Artificial Intelligence. When you drill down to the heart of our hardware and software designed to process images and video, what you find is years of study put in our developers. They’ve dug deeply into pattern recognition, their research run as joint projects between leading universities and Lanit-Tercom.
A scientific framework, top-notch professionalism, and a continual hunt for new ideas lets us create breakthrough technology, unique algorithms, and software without peer anywhere else in the world. No matter how complex the job, our team finds the best, smartest, and most cost-effective solution. Our portfolio includes projects for international clients as well as some of our own patented developments.
Services:
Security sensors for UAVs, motor vehicles, and robots
- Identifying static obstacles using stereo or neural networks, mobile obstacles using a video stream, the distance to the obstacle, road signs, and road irregularities
- Tracking objects located in video streams in real time
- Predicting trajectories of moving objects
- 360-degree vision systems
3D reconstruction
- Reconstructing scenes in motion from mobile or stereo cameras, shots from different angles
- Localizing cameras into already recorded scenes
- Tracking scene changes by comparing clouds
- Taking measurements in 3D scenes
Polarization
- Identifying surface defects using precise computations pulled from surface profiles
- Clarifying and filling out 3D models using surface profile calculations
Neural networks
- Facial and object recognition
- Developing brand new network architecture
Stitching panoramas
- 2D panorama stitching (360х180)
- Stitching 360-degree video panoramas for surveillance systems
Gesture recognition
- Gesture recognition in stereo
Support services
- High-precision calibration for the internal parameters used by cameras and camera systems
- Autocalibration for camera systems
- Automatic compensation for camera movement
- Pulling algorithms to different platforms (including built-in and mobile)