M4.31WN

General

The M4.31WN is a  sensor module that integrates deep learning, 3D computer vision and image processing functions. The M4.31WN is based on Inuitive’s NU4100AI multi-core processor SoC which is optimized for 3D image processing, computer vision and AI.

The power efficient M4.31WN is the leading solution for implementing 3D sensing and computer vision into consumer electronic devices. The M4.31WN is designed to function as a self-contained, selfsufficient depth sensor (at narrow Fov), vision and AI processor
module. It functions as an independent sub-system, generating fully processed depth information to its host.

The M4.31WN primary processing unit is Inuitive’s multi-core NU4100AI processor. In addition, the M4.31WN includes tracking and RGB sensors. These sensors are controlled and time stamped by the NU4100AI. Data generated by the sensors (and processed
in real-time by the NU4100AI) is used to generate a hybrid tracking based on 2D, 3D and IMU.

For design flexibility and enhanced performance, M4.31WN reference
design is comprised of two sub-modules:

• M4.31WN EOM (Electro Optical Module) – the EOM hosts the cameras, the projector and the IMU.
• M4.31WN Base – the base includes the NU4100AI IC, various interfaces, power-management circuit and other peripherals.

Key Features

• Active stereo depth sensor based on IR pattern illumination and IR monochrome sensors
  • 0.3-3 or 0.5-5 meters range (configurable settings)
• Narrow Field of View (65⁰x40⁰ FoV)
• RGB camera sensor, registered to the depth sensor
• 2 monochromatic Fish Eye cameras for tracking
• Utilizes Inuitive’s NU4100AI multi-core computer vision and 3D image processor
• On-board Interial Measurements Unit (IMU)
• Low power, small size
• Target Markets:
  • VR Head Mounted Displays (HMD)
  • AR/MR Glasses
  • Robots
  • Drones
• Enables AI processing on NU41000AI
• 3.2 TOPs Deep Learning processing
• Supports Inuitive AI framework
• Supports Windows, Linux, and Android OSs
• Operates in all lighting conditions (from darkness to daylight)
• API for easy Application-Layer Design
• On-chip Feature Tracking algorithms