Magic Eye Introduces Next-Generation ILT™ Evaluation Platform for Real-Time 3D Depth Sensing
High-density point cloud output at low latency enables realistic evaluation and rapid embedded system development
Magic Eye a developer of advanced 3D depth sensing technology based on ILT™ (Invertible Light Technology) , will unveil a new high-resolution, real-time ILT evaluation system at the upcoming Consumer Electronics Show (CES) designed to enable customers to evaluate ILT performance, validate configurations and begin application development for robotics, low-cost LiDAR-class replacement and automotive in-cabin applications.
This new evaluation system is a reference implementation, not a commercial sensor product. It outputs over 8,600 3D point clouds at 30 FPS, equivalent to over 259,000 depth points per second, while maintaining real-time operation and low latency (approximately 33 ms). This represents approximately double the spatial point density compared to Magic Eye's previous evaluation platform, without sacrificing frame rate.
"Customers evaluating depth sensing technology want realistic, real-time, and buildable data," said Skanda Visvanathan, vice president of business development at Magic Eye. "This reference system delivers high-resolution ILT depth at 30 FPS and is designed to shorten the evaluation to application development process in a form factor and performance range for embedded systems."
Designed for real-world evaluation and development, the evaluation system enables customers to evaluate ILT depth sensing in their own environment, begin application software development using live 3D point cloud output, and validate ILT configurations, including field of view, operating distance range, optical design, and processing pipeline, prior to custom module design.
Key characteristics of the evaluation platform include a wide viewing angle of 105° x 79°, a wide operating distance range of 0.3 to 2 m (encompassing close-range applications), and operation in bright indoor environments up to approximately 50,000 lux, depending on distance and target reflectivity.
Unlike depth solutions that reduce frame rate to increase point density, Magic Eye's ILT evaluation system maintains a full 30 FPS, enabling depth perception suited to changing real-time environments. With improved processing power, ILT™ can also operate at higher frame rates.
At CES, NVIDIA will demonstrate how the evaluation system supports development and prototyping for robotics applications such as real-time perception and navigation, low-cost LiDAR-class embedded sensing, and in-cabin occupant detection and interior monitoring in automobiles.
The evaluation system works with Magic Eye's MKE API, enabling customers to stream point clouds and integrate ILT depth data into their existing software stacks.
