The patented technology involves a multispectral automotive vision system consisting of a pair of stereoscopic infrared sensors and a pair of stereoscopic visible light sensors. The system includes a data fusion module that merges data received from infrared and visible light channels to enable accurate obstacle detection and distance estimation. The merger between the two stereoscopic channels also handles corner case scenarios, especially in difficult weather and lighting conditions, while reducing false alarms. The system’s automatic calibration module is designed to ensure that stereo cameras remain calibrated regardless of their configuration or position, to create an accurate and continuous perception of depth.

This patent serves as the underlying technology for the QuadSight automotive vision system, one of the company’s flagship products. Foresight’s four-camera vision system offers exceptional obstacle detection capabilities for semi-autonomous and autonomous vehicles. Through the fusion of sensors, QuadSight harnesses reflected light from visible light cameras with thermal energy captured by longwave infrared cameras for robust and accurate object detection of any shape, shape or material, in any environment. weather and lighting conditions – including total darkness, rain, haze, fog and glare.