hku-mars / FAST-LIVO

A Fast and Tightly-coupled Sparse-Direct LiDAR-Inertial-Visual Odometry (LIVO).
GNU General Public License v2.0
1.14k stars 181 forks source link

[laserMapping-1] process has died [pid 11250, exit code -11 #41

Closed IaroslavS closed 1 year ago

IaroslavS commented 1 year ago

If anybody faces this kind of error (look at the screen shot), please pay attention to string

printf("[ LIO ]: Using multi-processor, used core number: %s.\n", MP_PROC_NUM);

in laserMapping.cpp

Resolve the bug by changing the above string to the following:

printf("[ LIO ]: Using multi-processor, used core number: %d.\n", MP_PROC_NUM);

Screenshot from 2023-02-01 12-31-57

sebastiantorres1 commented 1 year ago

I had made a research in the lasermappin.cpp code and apparently the process stops working in this part of the code:

if(!vk::camera_loader::loadFromRosNs("laserMapping", lidar_selector->cam)) throw std::runtime_error("Camera model not correctly specified.");

Have you faced this problem?

I tried your solution and it doesn't fix my problem.

IaroslavS commented 1 year ago

vk::camera_loader::loadFromRosNs("laserMapping", lidar_selector->cam

Yes, I faced similiar problem, check out config file that you use when launching, for example this config file https://github.com/hku-mars/FAST-LIVO/blob/main/config/camera_NTU_VIRAL.yaml#L1 has field cam_model which should be one of the set {Ocam, Pinhole, ATAN}

Check or edit this file https://github.com/uzh-rpg/rpg_vikit/blob/master/vikit_ros/include/vikit/camera_loader.h that repository (https://github.com/uzh-rpg/rpg_vikit/) is a submodule for fast-livo

sebastiantorres1 commented 1 year ago

Apparently i just have to change the camera model "Pinhole" to "ATAN" in the configuration, but now it only shows the camera and the trajectory, but not the lidar odometry in the rviz like this image shows

Screenshot from 2023-02-06 10-07-11

How did you solve this?

xuankuzcr commented 1 year ago

As for our provided bags, you should use Pinhole camera model. While running your own collected dataset, choose the camera model according to your device.