OmniSense is a software platform developed based on automotive-grade deep learning algorithm, integrating modules of AI perception, event, early-warning, cloud monitoring, permissions. It supports fusion with external data and provides flexible and intelligent services and visualized interface for diverse scenarios such as V2X, smart highway, smart rail transit and industrial automation, etc.
OmniSense can detect when a pedestrian enters a pre-defined area, such as a crosswalk, or areas of the road outside of the crosswalk. This information is combined with input from the traffic light (green/yellow/red) to trigger an appropriate warning. This warning can be broadcast by a roadside unit (RSU) to nearby smart vehicles, or displayed in a prominent location to alert oncoming drivers.
OmniSense can be configured so when abnormal events occur, the point cloud is saved as a 3D point cloud or a 2D image format. The data recorded in real time can be sent back to a server for storage, where it can be accessed according to time stamp.
Due to the long-range detection capability of Innovusion LiDAR, OmniSense can detect events over 200 meters away, such as pedestrians crossing the road, and issue an appropriate warning. Unlike other LiDARs which are limited to sensing events within 50-100 meters, OmniSense can truly observe hazards beyond the “horizon” typically noticed by drivers. Event warnings can be transmitted by roadside unit and broadcast to nearby vehicles.
Blind spot scenarios, such as pedestrians obscured by larger vehicles, can be detected by the 360° Multi LiDAR Fusion. A warning can be broadcast to other vehicles through roadside units.
Object Detection
Object Classification
Continuous Calibration
Analyzed Data Output
Flexible Data Access
Multi-LiDAR Fusion
The product you are interested in : OmniSense
