Melexis and emotion3D have worked together to integrate DMS and HUD dynamic object correction in a single camera

     Melexis and emotion3D, a leading provider of camera-based in-car analytics software, have launched a 3D time-of-flight (ToF) sense-based driver Monitoring System (DMS) demo that, combined with high-precision 3D eye positioning, enables augmented reality head-up display (AR HUD) dynamic correction.

     Melexis and emotion3D's new DMS module covers all basic DMS functions, such as driver sleepiness and attention warning, in perfect compliance with Europe's common safety regulations and EuroNCAP's new test standards. In addition, the demonstration device provides 3D positioning of the driver's facial features, such as accurate head orientation recognition and eye positioning. These are the prerequisites for ensuring the best possible experience with an augmented reality head-up display (AR HUD). Objects projected by the HUD need to be precisely corrected for real-world objects and dynamically change with the position of the driver's head and eyes.

     The demonstration unit consists of Melexis' 3D ToF sensor chip MLX75027 and a camera built by emotion3D's Advanced In-Car Analytics software (E3D ICMS). The software can be flexibly integrated into a specified vehicle scale SoC.

     "Due to regulatory requirements, the integration of DMS into new models is imperative, and augmented reality head-up displays are also gaining popularity. Our system brings the two together to provide a highly accurate and cost-effective solution for car manufacturers, "said Florian Seitner, CEO of emotion3D.

     Gualtiero Bagnuoli, Melexis product Marketing Manager, commented on the benefits of ToF technology: "We combine accurate and reliable 3D eye position detection for HUD with sun-free line of sight and eye opening and closing detection to enable a more streamlined DMS algorithm. Among them, ToF technology is crucial and can effectively improve work efficiency. Getting accurate depth data from the 3D ToF sensor chip is very easy and the processing effort is very low. The DMS and HUD algorithms work seamlessly with the wide-field lenses and VGA-resolution ToF sensor chips."