top of page

ZED 2

From sensors to software, the ZED 2 camera is filled with cutting-edge technology that takes spatial perception to a whole new level.

 

Neural Depth Sensing

ZED 2 is the first stereo camera that uses neural networks to reproduce human vision, bringing stereo perception to a new level.

 

Built-in IMU, Barometer & Magnetometer

Multi-sensor capture made easy. Gather real-time synchronized inertial, elevation and magnetic field data along image and depth.

 

120° Wide-Angle Field of View

With its 16:9 native sensors and ultra sharp 8-element all glass lenses, capture video and depth with up to 120° field of view.

 

Spatial Object Detection

Detect objects with spatial context. Combine AI with 3D localization to create next-generation spatial awareness.

 

Improved Positional Tracking

Benefit from a wide angle FOV, advanced sensor stack and thermal calibration for a greatly improved positional tracking accuracy.

 

Cloud Connected

Monitor and control your camera remotely. Using a dedicated cloud platform, capture and analyze 3D data anywhere in the world.

 

This new stereo camera combines advanced depth sensing with AI, letting you build cutting-edge spatial intelligence and virtually replicate the physical world.

 

An all-new lightweight neural network for stereo matching brings stereo depth sensing to the next level, with a wide 110° x 70° FOV.

 

 

VIDEO OUTPUT

Video ModeFrames per secondOutput Resolution (side by side)
2.2K154416x1242
1080p30 / 153840x1080
720p60 / 30 / 152560x720
WVGA100 / 60 / 30 / 151344x376

 

Video Recording

Native resolution video encoding in H.264, H.265
or lossless format (on host)

Video Streaming

Stream anywhere over IP using ZED SDK

ISP

New ISP tuned with machine learning for AI and vision tasks

 

DEPTH

Depth Resolution

Native video resolution
(in Ultra mode)

Depth FPS

Up to 100Hz

Depth Range

0.2 - 20 m (0.65 to 65 ft)

Depth FOV

110° (H) x 70° (V) x 120° (D) max.

Technology

Neural Stereo Depth Sensing
 

MOTION

Motion Sensors

Accelerometer
Gyroscope
Data Rate: 400Hz

Pose Update Rate

Up to 100Hz

Position Sensors

Barometer
Magnetometer
Data Rate: 25Hz / 50Hz

Pose Drift

Translation: 0.35%
Rotation: 0.005°/m
(without loop correction)
Technology
6-DoF visual-inertial stereo SLAM with advanced sensor fusion and thermal compensation


OBJECT DETECTION
Object Types
Persons, Vehicles

Object Tracking
Yes
Detection Outputs
Bounding Boxes 2D/3D

Location

Speed

Unique ID

Segmentation Masks
Detection Range
Up to 20m (3D)
Up to 40m (2D)
Skeleton Tracking
18x Body Keypoints
Head Bounding Box
Up to 10m Range

 

LENS

Lens Type

Wide-angle 8-element all-glass dual lens with optically corrected distortion

Aperture

ƒ/1.8

Field of View

110° (H) x 70° (V) x 120° (D) max.

 

IMAGE SENSORS

Sensor Resolution

Dual 4M pixels sensors with 2-micron pixels

Sensor Format

Native 16:9 format for a greater horizontal field of view

Sensor Size

1/3” backside illumination sensors with high low-light sensitivity

Shutter Sync

Electronic Synchronized

Rolling Shutter

Camera Controls

Adjust Resolution, Frame rate,

Brightness, Contrast,

Saturation, Gamma,

Sharpness, Exposure and

White Balance

 

SENSORS
Accelerometer

Gyroscope

Barometer

Magnetometer

Temperature sensor

 

CONNECTIVITY

Connector

USB 3.0 port with 1.2m integrated cable

Mounting Options

1/4"-20 UNC thread mount
2x M3 threads

Power

Power via USB 5V / 380mA

Operating Temperature

-10°C to +50°C (14°F to 122°F)

 

SIZE & WEIGHT

Dimensions

175 x 30 x 33 mm (6.89 x 1.18 x 1.3’’)

Weight

124 g (0.35 lb)

Material

All-aluminium enclosure with thermal compensation of factory calibration

COMPATIBLE OS

Windows 10, 8, 7

Ubuntu 18, 16

Debian, CentOS (via Docker)

Jetson L4T

 

SDK SYSTEM REQUIREMENTS

Dual-core 2.3GHz or faster processor

4 GB RAM or more

Nvidia GPU with compute capability > 3.0

 

IN THE BOX

ZED 2 Stereo camera

Mini Tripod stand

Documentation


10-year hardware warranty for all ZED 2 cameras

 

DOWNLOAD STEREOLABS ZED 2 CAMERA PDF

Get in Touch
bottom of page