https://github.com/ecmnet/mavslam

Visual odometry based on Intel® RealSense™ devices
https://github.com/ecmnet/mavslam

computervision obstacle-avoidance odometry px4 realsense

Last synced: about 1 year ago
JSON representation

Visual odometry based on Intel® RealSense™ devices

• Host: GitHub
• URL: https://github.com/ecmnet/mavslam
• Owner: ecmnet
• Created: 2016-03-12T14:44:03.000Z (about 10 years ago)
• Default Branch: master
• Last Pushed: 2019-10-24T14:49:16.000Z (over 6 years ago)
• Last Synced: 2025-03-26T12:51:12.997Z (about 1 year ago)
• Topics: computervision, obstacle-avoidance, odometry, px4, realsense
• Language: Java
• Homepage:
• Size: 37.7 MB
• Stars: 51
• Watchers: 5
• Forks: 12
• Open Issues: 4
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# MAVSlam - Visual odometry & SLAM for PX4

## 

#### Onboard visual odometry for PX4 

(Deprecated)

Integrate  Intel® RealSense™ R200 depth camera as an oboard visual input source for PX4 using the [UP-Board]( http://www.up-board.org) as companion:

- Estimate position and speed based on stereo depth images up to 60 fps for LPE /EKF2

- Stream video to MAVGCL with overlays 

- Feature-based or depth obstacle detection

- Integration with PX4 via MAVLink ([MAVComm](https://github.com/ecmnet/MAVComm) required)

Prerequisites:

- UP-Board running Ubilinux 3.0 with Java 8 (minimum) stack connected via serial link to PX4 controller

- Intel® RealSense™ R200 (connected with external power supply)

Video:

Heading estimation:

![https://raw.githubusercontent.com/ecmnet/MAVSlam/master/MAVSlam/heading.png](https://raw.githubusercontent.com/ecmnet/MAVSlam/master/MAVSlam/heading.png)

XY Estimation:

![https://raw.githubusercontent.com/ecmnet/MAVSlam/master/MAVSlam/screenshot6.png](https://raw.githubusercontent.com/ecmnet/MAVSlam/master/MAVSlam/screenshot6.png)

Obstacle detection:

![https://raw.githubusercontent.com/ecmnet/MAVSlam/master/MAVSlam/obstacle.png](https://raw.githubusercontent.com/ecmnet/MAVSlam/master/MAVSlam/obstacle.png)

MicroSLAM:

![https://raw.githubusercontent.com/ecmnet/MAVSlam/master/MAVSlam/microslam.png](https://raw.githubusercontent.com/ecmnet/MAVSlam/master/MAVSlam/microslam.png)

#### Desktop odometry demo 

A first implementation using Intel® RealSense™ R200 with boofcv on OSX/Linux platforms. Based on https://github.com/IntelRealSense/librealsense and http://boofcv.org/index.php?title=Main_Page.

![https://raw.githubusercontent.com/ecmnet/MAVSlam/master/MAVSlam/RealSense.png](https://raw.githubusercontent.com/ecmnet/MAVSlam/master/MAVSlam/RealSense.png)