RGB ToF 3D Solutions

Product Introduction

We provide a variety of RGB ToF 3D solutions for different application scenarios.

Spot iToF/Spot dToF depth completion solution: Constructs high-integrity, low-noise depth images through the fusion of RGB and sparse depth. This solution delivers a 15dB improvement in SNR, a 100x improvement in depth image resolution, and vast improvements in depth measurement distance and MPI. This provides complete depth information for video bokeh, 3D mesh, AR, photograph, and other applications and constructs the third dimension of the metaverse.

iToF AI depth-perception solution: In the ToF RAW domain, the AI probability-regression model removes the inherent noise of the sensor and the noise introduced by shooting distance and scene materials, which allows for improved retention of depth accuracy and details. Application areas: face verification unlock, gesture recognition, etc.

dToF single-photon on-chip solution: Solves problems with high storage and bandwidth requirements, fully extracts data information, and reduces TDC sampling rate demands. Application areas: robots, AR, vehicle-mounted lidar, etc.

Supported Platform

Qualcomm 8 Series, MTK

Product Advantages

Tetras.AI's RGBToF solution constructs RGBD images with high integrity and low noise. The solution improves SNR by 15dB, while improving depth image resolution 100x, and vastly improving depth measurement distance and MPI. In addition, our RGBToF solution offers advantages such as excellent real-time performance (supports 30FPS video) and low power consumption (lowest power consumption <230MW)

Key Technologies

ToF Depth Engine

Uses depth acquisition channels such as binocular, monocular, and ToF to intelligently select appropriate algorithm combinations (including AI stereo matching, AI monocular depth, AI-ToF, and AI RGBD fusion) for depth estimation.
AI Depth Noise-Reduction Model

In the ToF RAW domain, the AI probability-regression model removes the inherent noise of the sensor and the noise introduced by shooting distance and scene materials to improve the retention of depth accuracy and details.
RGBToF Fusion Technology

Intelligently judges scene features, fuses RGB scene information, and completes the missing depths, including completing spot depths and missing depths over a large area.

Application Scenarios

AR
Face Verification