Autonomous machines and vehicles rely on different sensors to create an accurate 3D perception of the environment. Stereoscopic vision systems are composed of two cameras that must remain calibrated at all times in order to ensure the accuracy of the 3D image they generate.

Stereo cameras are directly impacted by minor changes, such as slight movements, vibrations and temperature differences, which can cause a reduction in accuracy and sometimes even a complete loss of calibration. Our software-based solution enables continuous calibration between pairs of vision sensors, including thermal infrared and visible-light cameras, and can enhance existing stereo systems.

.