RGB-D Codec

Realtime 3D video streaming

High resolution depth maps

Fast video compression

Realtime 3D video streaming
High resolution depth maps
Fast video compression

3d video is bandwidth intensive

Automating tasks in complex scenarios requires excellent knowledge of the environment, placing high demands on sensor hardware and sensor processing. High-resolution depth cameras that serve this need have become a commodity. High-resolution data, especially depth data requires significant computation power to process the data. At about 480 Mbps (uncompressed depth video, 720p, 30fps) – available bandwidth, as well as the lack of hardware-accelerated compression algorithms that retain sufficient depth and RGB information has limited streaming solutions so far.

Automating tasks in complex scenarios requires excellent knowledge of the environment, placing high demands on sensor hardware and sensor processing. High-resolution depth cameras that serve this need have become a commodity. High-resolution data, especially depth data requires significant computation power to process the data. At about 480 Mbps (uncompressed depth video, 720p, 30fps) – available bandwidth, as well as the lack of hardware-accelerated compression algorithms that retain sufficient depth and RGB information has limited streaming solutions so far.

Read more

The aivero 3dQ codec

A C++ library for efficiently compressing high-resolution Depth and RGB video streams using hardware accelerated H265 codecs. Retain high-quality depth data, ready for even the most demanding development projects, while significantly reducing the bandwidth required to transmit the video or space required to store it.

A C++ library for efficiently compressing high-resolution Depth and RGB video streams using hardware accelerated H265 codecs. Retain high-quality depth data, ready for even the most demanding development projects, while significantly reducing the bandwidth required to transmit the video or space required to store it.

Read more

3dq Performance

Here you can compare the raw uncompressed RGB-D video with with the Aivero 3DQ codec. Even after reducing the size of the video by over 95% we are able to provide a high quality video stream that supports high demand machine vision applications.

coverup-wide

Applications

Industrial Robotics
Drones

Autonomous Vehicles

IOT

Machine Vision

CCTV

camera agnostic

Alongside the pixels of the regular colour or grey-scale image, depth cameras capture the distance to the camera. Depth perception enables the system to precisely measure, differentiate and classify objects. Different cameras use various techniques to measure distance such as stereo vision, projected lighting and time of flight. Driven by the recent development of autonomous cars, depth cameras have significantly improved in recent years and will likely continue to do so.

Alongside the pixels of the regular colour or grey-scale image, depth cameras capture the distance to the camera. Depth perception enables the system to precisely measure, differentiate and classify objects. Different cameras use various techniques to measure distance such as stereo vision, projected lighting and time of flight. Driven by the recent development of autonomous cars, depth cameras have significantly improved in recent years and will likely continue to do so.

Read more

SPECIFICATIONS

3dq codec:

Range:20 – 100 cm (extended range version TBA)
Resolution:Up to 1080p
Frame rate:Up to 30 fps
Supported OS:Linux

3DQ Streaming Protocol:

Transport:WebRTC and RTP
Latency:100 ms
Encryption:Up to 30 fps
Jitter buffering:Configurable