In Progress

ROS Recording, Gazebo Simulation Data II

Purpose of this project:

Create a sample data set from the perspective of an autonomous car in simulation.

Usually autonomous cars have lidar and camera sensors. Gazebo can simulate both of them and there is a prebuilt city, vehicles and sensors available. In order to run Gazebo in a usable way you need a _powerful_ machine.

Steps:

1. Install ROS (robot operating system) melodic (releasename)in the "full-desktop"-version on ubuntu bionic beaver 18.04

2. Install mercurial source control management (sudo apt-get install hg)

3. Clone repository and build [login to view URL] (hg clone [login to view URL] )

** How to build is noted on the repository page

--> this should give you the environment and a vehicle with a lidar sensor

4. Add 8 cameras around the vehicle, please note the position, direction and parameters of the cameras to me.

5. Record ros-bag file from the moving vehicle in the city

* 30 seconds should be enough

* manually drive the vehicle through simulation

* make sure the vehicle is turning (not only going straight) and there are other cars visible within the sensors field of view

* make sure different kinds of buildings are visible in the sensors field of view

* Cameras and lidar shall gather data synchronized (not more than 1ms apart in simulation time, better zero)

* Also record the absolute position of the autonomous vehicle

* Drive by atleast one traffic light

6. From this ros-bag the following data for each time step

* .png images of all the cameras

* save the pointcloud as a binary file with the extension .bin

** using this function [login to view URL]

** pointcloud must be in 3D (x,y,z)

** pointcloud datatype shall be float32

Deliver the following artifacts

** Position and Orientation of each camera relative to the vehicle

** Position and Orientation of the lidar sensor relative to the vehicle

** The recorded ros-bag file

** The extracted images and pointclouds for each timestep of the ros-bag file as a zipped folder (this is the most important delivery)

** The script to extract the data from the ros-bag file

** The changes to citysim in order to get the cameras running

Budget is 50 Euros. If you have worked with ROS and Gazebo before this should be doable in less than 2 days.

Taidot: Python, Robotiikka

Näytä lisää: data entry cost per record canada, data entry cost per record, simulation data real time ruby, gazebo record video, ros video capture, ros record video, data entry bid per record, simulation data structure cashier, simulation data matlab nntool, simulation multisim traffic light, traffic light simulation, recording stock exchange data real time, traffic light simulation online, google analytics simulation data, simulation data link layer, simulation data google analytics, analyse raw data and come up with useful insights trends or warnings from the data, data entry services offered for a wide range of projects from mailing lists to manuscripts, mobile/tablet data collection software needed like surveytogo with offline and online data collection purpose, mobiletablet data collection software needed like surveytogo with offline and online data collection purpose

Tietoa työnantajasta:
( 1 arvostelu ) Chemnitz, Germany

Projektin tunnus: #17545246

Myönnetty käyttäjälle:

ebadsyed

bid bid bid bid .....................................................................

€55 EUR 3 päivässä
(0 Arvostelua)
0.0

2 freelanceria on tarjonnut keskimäärin %project_bid_stats_avg_sub_26% %project_currencyDetails_sign_sub_27% tähän työhön

naishodayo

Hi, I am an experienced programmer and I'm sure I can complete your project asap. Please contact me and tell about your project details. I promise a high quality and punctual work. Thanks. ROS Recording, Gazebo Simula Lisää

€155 EUR 1 päivässä
(0 arvostelua)
0.0