GPU & Compute

DVO: dense visual odometry

Real-time SLAM system for RGB-D cameras using dense stereo matching.

Status
UnmaintainedNo commits in 4719 days.
Last checked
Checked · Oct 8, 2026
License
Open source
Pricing
Free tier · from $0
Visit DVO: dense visual odometry Compare with alternatives

What it is

Dense Visual Odometry (DVO) is a real-time simultaneous localization and mapping (SLAM) system designed to work with RGB-D cameras. It uses dense stereo matching to estimate camera motion and build a map of the environment, making it valuable for robotics and augmented reality applications.

DVO stands out by offering high-precision dense stereo matching in a self-hosted, open-source library format, making it ideal for developers who need robust SLAM capabilities without cloud dependencies.

When it fits — and when it doesn’t

A good fit for

  • Developers working on real-time robotics projects requiring precise motion estimation

  • Research teams focusing on indoor/outdoor mapping and localization using RGB-D cameras

  • AR developers needing a robust SLAM system for accurate spatial understanding

Look elsewhere if

  • Projects that require real-time processing without access to powerful hardware

  • Applications where the use of C++ is not feasible or preferred

Known weaknesses

Each one links to the public issue, changelog or docs page it came from, where we have one.

  1. 01

    Steep learning curve for developers unfamiliar with C++

    DVO's implementation in C++ requires a strong understanding of the language, which can be challenging for those more accustomed to higher-level languages.

  2. 02

    Limited support and community size compared to larger SLAM libraries

    The open-source nature of DVO means that contributions and updates are less frequent than in more popular frameworks, potentially leading to slower issue resolution and feature development.

  3. 03

    Performance issues with high-resolution RGB-D cameras

    Dense stereo matching can become computationally expensive with higher resolution input data, causing real-time performance degradation on less powerful hardware.

  4. 04

    Limited flexibility in customization and extension

    The architecture of DVO is tightly coupled, making it difficult to integrate custom modules or extend functionality without significant modifications to the core codebase.

What it does well

  • Real-time SLAM capabilities for RGB-D cameras

  • Uses dense stereo matching for accurate motion estimation

  • Suitable for both indoor and outdoor environments

  • Highly customizable with extensive documentation

Pricing

Free tier
YesOpen source — free to use
Paid from
$0
Model
Flat rate
Enterprise plan
No

Pricing not independently confirmed — check the vendor’s page.

Performance benchmarks

How Fast Is It?

Set up DVO: dense visual odometry

Setup steps, common gotchas and what to check after 30 days.

Open the setup guide →