isee-technology/camvox

[ICRA2021] A low-cost SLAM system based on camera and Livox lidar.

camvox's Issues

Lidar-Camera-Calibration

作者您好，我查看标定代码CamVox/isee-camvox/src/Calibrating.cc发现并没有使用关于位姿的优化算法，而是使用了类似CSM的位姿暴力搜索算法，这和论文中关于Lidar-Camera标定配图的ICP迭代不太相符？

rosrun 运行问题

我们看到您团队的CamVox项目的实现效果非常不错，对这一项目非常感兴趣。
感谢您在github上将其开源，我们也试着将其复现。但是我在安装运行过程中，遇到了一些问题无法解决，希望请教一下

我按照步骤安装，但是没装livox-sdk与mvs camera-sdk。我想只运行一下rosbag看看效果。

我打开了roscore ，之后运行 rosrun online camvox Vocabulary/ORBvoc.bin camvox/online/Livox.yaml 时，出现segement failed。

经过调试，我发现错误出现在calibrating.cc文件中，cv::projectPoints(pts_3d, r_vec, t_vec, camera_matrix, distortion_coeff, pts_2d);这个函数出错，
pts_3d, r_vec, t_vec, camera_matrix, distortion_coeff等参数均有对应数值，但是输出的pts_2d值为nan，导致之后的程序无法运行。

由于在rosrun这一步已经出错，我没有运行rosbag。

感谢您的帮助

missing of MvErrorDefine.h

Thanks for your work.
When I tried compiling CamVox, it gave me several errors, one is

/catkin_ws/src/CamVox-main/isee-preprocessing/src/isee_synchronize.cpp:3:10: fatal error: isee_synchronize/CustomMsg.h: No such file or directory
 #include <isee_synchronize/CustomMsg.h>

I found the required file at
catkin_ws/src/CamVox-main/isee-preprocessing/build/devel/include/isee_synchronize
Thus I copy the required file to include dir, and it solved.
Another error is

/catkin_ws/src/CamVox-main/isee-preprocessing/include/camera.h:8:10: fatal error: MvErrorDefine.h: No such file or directory
 #include "MvErrorDefine.h"

I checked my workspace and there is no such file, then I searched online and found it should be a Halcon camera file, but I dont know which file you are using.

GPS Hardware

Hi, which GPS hardware are you using ?

fatal error: pcl/common/transforms.h: 没有那个文件或目录

./build_ros.sh时报错，请问怎么解决

Livox Avia

Hi ,is it possible use livox Avia with it?

Thanks by share

How can I use r3live rosbag ?

thanks to release this excelent project ! I want to compare camvox with r3live, but I think I should generate isee-depth image. I find livox_lidar2depth fuction in isee_synchronize.cpp,but I don't know how to set the input paramters of createFromPointCloudWithFixedSize. I hope to get your help,thanks !

Does it work on other Livox LiDAR models?

Great work! We are planning to adapt this algorithm to other Livox LiDAR-camera combinations, so does this algorithm generalize to other LiDAR models, i.e., MID-40/70/100 or TELE series yet?

Runtime error : Segmentation fault (core dumped)

Hello,Thank you for your great work!
When I run the project with Rosbag Example .It runs normally at first, But a Segmentation fault occurs after a few minutes.
The following is the log print information，But I can't find out what went wrong. looking forward to your reply !

Camvox Copyright (C) 2020 ISEE, University of SUSTech.
This program comes with ABSOLUTELY NO WARRANTY;
This is free software, and you are welcome to redistribute it
under certain conditions. See LICENSE.txt.

Input sensor was set to: RGB-D

Loading ORB Vocabulary. This could take a while...
Vocabulary loaded in 0.31s

Camera Parameters:

• fx: 957.994
• fy: 955.328
• cx: 790.335
• cy: 250.663
• k1: -0.12
• k2: 0.1162
• p1: 0
• p2: 0
• fps: 10
• color order: RGB (ignored if grayscale)

ORB Extractor Parameters:

• Number of Features: 1500
• Scale Levels: 8
• Scale Factor: 1.2
• Initial Fast Threshold: 20
• Minimum Fast Threshold: 7

Depth Threshold (Close/Far Points): 129.998
Framebuffer with requested attributes not available. Using available framebuffer. You may see visual artifacts.Robot keep 1s still !
Robot keep 1s still !
Robot keep 1s still !
Robot keep 1s still !
Robot keep 1s still !
Calibrating processing
Robot keep 1s still !
New map created with 159 points
Calibrating processing
cancel calibration !
Calibrating RELEASE
[pcl::VoxelGrid::applyFilter] Leaf size is too small for the input dataset. Integer indices would overflow.After spin:

Saving keyframe trajectory to OnlineKeyFrameTrajectory.txt ...

trajectory saved!
Robot keep 1s still !
Dense Color Pointcloud save finished
QObject::QObject: Timers cannot be stopped from another thread
Segmentation fault (core dumped)
(base) lqql@lqql:/catkin_ws/src/CamVox/isee-camvox$ rosrun online camvox Vocabulary/ORBvoc.bin camvox/offline/Livox.yaml 1000

Camvox Copyright (C) 2020 ISEE, University of SUSTech.
This program comes with ABSOLUTELY NO WARRANTY;
This is free software, and you are welcome to redistribute it
under certain conditions. See LICENSE.txt.

Input sensor was set to: RGB-D

Loading ORB Vocabulary. This could take a while...
Vocabulary loaded in 0.33s

Camera Parameters:

• fx: 957.994
• fy: 955.328
• cx: 790.335
• cy: 250.663
• k1: -0.12
• k2: 0.1162
• p1: 0
• p2: 0
• fps: 10
• color order: RGB (ignored if grayscale)

ORB Extractor Parameters:

• Number of Features: 1500
• Scale Levels: 8
• Scale Factor: 1.2
• Initial Fast Threshold: 20
• Minimum Fast Threshold: 7

Depth Threshold (Close/Far Points): 129.998
Framebuffer with requested attributes not available. Using available framebuffer. You may see visual artifacts.New map created with 133 points
After spin:

Saving keyframe trajectory to OnlineKeyFrameTrajectory.txt ...

trajectory saved!
Robot keep 1s still !
Dense Color Pointcloud save finished
QObject::QObject: Timers cannot be stopped from another thread
Segmentation fault (core dumped)
(base) lqql@lqql:/catkin_ws/src/CamVox/isee-camvox$ rosrun online camvox Vocabulary/ORBvoc.bin camvox/offline/Livox.yaml 1000

Camvox Copyright (C) 2020 ISEE, University of SUSTech.
This program comes with ABSOLUTELY NO WARRANTY;
This is free software, and you are welcome to redistribute it
under certain conditions. See LICENSE.txt.

Input sensor was set to: RGB-D

Loading ORB Vocabulary. This could take a while...
Vocabulary loaded in 0.32s

Camera Parameters:

• fx: 957.994
• fy: 955.328
• cx: 790.335
• cy: 250.663
• k1: -0.12
• k2: 0.1162
• p1: 0
• p2: 0
• fps: 10
• color order: RGB (ignored if grayscale)

ORB Extractor Parameters:

• Number of Features: 1500
• Scale Levels: 8
• Scale Factor: 1.2
• Initial Fast Threshold: 20
• Minimum Fast Threshold: 7

Depth Threshold (Close/Far Points): 129.998
Framebuffer with requested attributes not available. Using available framebuffer. You may see visual artifacts.New map created with 87 points
Track lost soon after initialisation, reseting...
System Reseting
Reseting Local Mapper... done
Reseting Loop Closing... done
Reseting Database... done
New map created with 105 points
Track lost soon after initialisation, reseting...
System Reseting
Reseting Local Mapper... done
Reseting Loop Closing... done
Reseting Database... done
New map created with 87 points
Track lost soon after initialisation, reseting...
System Reseting
Reseting Local Mapper... done
Reseting Loop Closing... done
Reseting Database... done
New map created with 86 points
Track lost soon after initialisation, reseting...
System Reseting
Reseting Local Mapper... done
Reseting Loop Closing... done
Reseting Database... done
New map created with 79 points
Track lost soon after initialisation, reseting...
System Reseting
Reseting Local Mapper... done
Reseting Loop Closing... done
Reseting Database... done
New map created with 89 points
[pcl::VoxelGrid::applyFilter] Leaf size is too small for the input dataset. Integer indices would overflow.After spin:
Robot keep 1s still !
[pcl::VoxelGrid::applyFilter] Leaf size is too small for the input dataset. Integer indices would overflow.
Saving keyframe trajectory to OnlineKeyFrameTrajectory.txt ...

trajectory saved!
Dense Color Pointcloud save finished
Segmentation fault (core dumped)

problem with build 'build_ros.sh' in Ubuntu 18.04 ROS melodic

I get a error code like this, how can i solve it?

/usr/include/pcl-1.8/pcl/visualization/point_cloud_geometry_handlers.h:48:10: fatal error: vtkSmartPointer.h: No such file or directory
#include <vtkSmartPointer.h>
^~~~~~~~~~~~~~~~~~~
compilation terminated.
CMakeFiles/camvox.dir/build.make:118: recipe for target 'CMakeFiles/camvox.dir/src/ros_rgbd.cc.o' failed
make[2]: *** [CMakeFiles/camvox.dir/src/ros_rgbd.cc.o] Error 1
CMakeFiles/Makefile2:67: recipe for target 'CMakeFiles/camvox.dir/all' failed
make[1]: *** [CMakeFiles/camvox.dir/all] Error 2
Makefile:129: recipe for target 'all' failed
make: *** [all] Error 2

Sensor intrinsic and extrinsic

Thank you for sharing the code.
I read the code and found that the sensor extrinsic and intrinsic parameters are unclear.
The camera intrinsic in the code are different from those in yaml. #25 mentioned that the camera was changed once. So which is the accurate intrinsic?
In addition, it is mentioned in #7 that the imu information is not obtained from horizon, but from Inertial Sense uINS. So what is the extrinsic from imu to horizon or imu to camera?
So could you provide a yaml file corresponding to the Camvox.bag dataset, including the intrinsic and extrinsic of imu-camera-livox.
Thanks again, sincerely.

关于imu

您好，请问在这个CamVox解决方案里imu数据是从horizon激光雷达内置的imu芯片直接读取的还是从Inertial Sense uINS里读取出来的？

MV-CE060-10UC ROS driver?

Is there a ros driver compatible with MV-CE060-10UC?
I cant seem to get it working

thanks your work,but meet error Couldn't find executable named camvox,when I execute rosrun online camvox Vocabulary/ORBvoc.bin camvox/offline/Livox.yaml 1000

when I input:rosrun online camvox Vocabulary/ORBvoc.bin camvox/offline/Livox.yaml 1000
error:[rosrun] Couldn't find executable named camvox below /home/zehao/catkin_ws/src/CamVox/isee-camvox/camvox/online

Cannot download data

Hi!
I cannot download data from here http://192.168.3.3:20441/index.php/s/bj55H4bgYh2bqSF and from here http://www.zywok.com:20441/index.php/s/c3vmOIpaSMjKbD7?
New tab opens and page loaded infinitely
Also that is true for all links on zywok.com domen

Is there way to upload, at least, CamVox.bag to Google Drive, for example?

Trajectory.txt

请问下是怎么输出轨迹文件KeyFrameTrajectory.txt的？谢谢

Do you have a rosbg download address or baiduyunpan address

Do you have a rosbg download address or baiduyunpan address

why not use imu linear_acceleration to calculate translation for undistortion

It looks like now this repo only use rotation info from imu.
why not use imu linear_acceleration to calculate translation for undistortion?

Most links to the suggested prerequisites are mixed

They all link to a eigen tar file, can you fix it?

The dataset in SUSTech campus with loop closure does not contain lidar data

The bag only contains rgb image and depth image, but I hope to learn how to undistorted point cloud using imu data and project the lidar point cloud to image.
will this part of code be released later？

Hard Synchronization

I am very interested in your work. We are doing hardware time synchronization. You should be doing this part of the work through STM32. Can you provide this part of the code?

Build Error

Thanks for your work! I had trouble when build the code with ./build_ros.sh
It seems like something wrong with vtk, pcl and boost.
I reinstalled pcl1.7 and vtk6.2 via source code, but it didn't help.
Also, could you provide the way to build "libCamvox.so"? In that case I can test the code with other machines.
Thanks!!!

/usr/bin/ld: warning: libboost_system.so.1.65.1, needed by /opt/ros/melodic/lib/libroscpp.so, may conflict with libboost_system.so.1.58.0
/usr/bin/ld: warning: libboost_filesystem.so.1.65.1, needed by /opt/ros/melodic/lib/libroscpp.so, may conflict with libboost_filesystem.so.1.58.0
/usr/bin/ld: warning: libboost_thread.so.1.58.0, needed by ../../../lib/libCamvox.so, not found (try using -rpath or -rpath-link)
/usr/bin/ld: warning: libvtkRenderingFreeTypeOpenGL-6.2.so.6.2, needed by ../../../lib/libCamvox.so, not found (try using -rpath or -rpath-link)
/usr/bin/ld: warning: libvtkRenderingMatplotlib-6.2.so.6.2, needed by ../../../lib/libCamvox.so, not found (try using -rpath or -rpath-link)
/usr/bin/ld: warning: libvtkRenderingFreeTypeFontConfig-6.2.so.6.2, needed by ../../../lib/libCamvox.so, not found (try using -rpath or -rpath-link)
/usr/bin/ld: warning: libopencv_calib3d3.so.3.3, needed by ../../../lib/libCamvox.so, not found (try using -rpath or -rpath-link)
/usr/bin/ld: warning: libopencv_features2d3.so.3.3, needed by ../../../lib/libCamvox.so, not found (try using -rpath or -rpath-link)
/usr/bin/ld: warning: libopencv_highgui3.so.3.3, needed by ../../../lib/libCamvox.so, not found (try using -rpath or -rpath-link)
/usr/bin/ld: warning: libopencv_imgcodecs3.so.3.3, needed by ../../../lib/libCamvox.so, not found (try using -rpath or -rpath-link)
/usr/bin/ld: warning: libopencv_imgproc3.so.3.3, needed by ../../../lib/libCamvox.so, not found (try using -rpath or -rpath-link)
/usr/bin/ld: warning: libGLEW.so.1.13, needed by ../../../lib/libCamvox.so, not found (try using -rpath or -rpath-link)
/usr/bin/ld: warning: libvtkInteractionStyle-6.2.so.6.2, needed by ../../../lib/libCamvox.so, not found (try using -rpath or -rpath-link)
/usr/bin/ld: warning: libvtkRenderingOpenGL-6.2.so.6.2, needed by ../../../lib/libCamvox.so, not found (try using -rpath or -rpath-link)
/usr/bin/ld: warning: libvtkRenderingFreeType-6.2.so.6.2, needed by ../../../lib/libCamvox.so, not found (try using -rpath or -rpath-link)
/usr/bin/ld: warning: libvtkRenderingCore-6.2.so.6.2, needed by ../../../lib/libCamvox.so, not found (try using -rpath or -rpath-link)
/usr/bin/ld: warning: libvtkCommonDataModel-6.2.so.6.2, needed by ../../../lib/libCamvox.so, not found (try using -rpath or -rpath-link)
/usr/bin/ld: warning: libvtkCommonMath-6.2.so.6.2, needed by ../../../lib/libCamvox.so, not found (try using -rpath or -rpath-link)
/usr/bin/ld: warning: libvtkCommonCore-6.2.so.6.2, needed by ../../../lib/libCamvox.so, not found (try using -rpath or -rpath-link)
../../../lib/libCamvox.so：对‘vtkRenderingFreeTypeOpenGL_AutoInit_Destruct()’未定义的引用
../../../lib/libCamvox.so：对‘pcl::search::Searchpcl::PointXYZ::getName() const’未定义的引用
../../../lib/libCamvox.so：对‘boost::this_thread::hiden::sleep_until(timespec const&)’未定义的引用
../../../lib/libCamvox.so：对‘vtkRenderingFreeTypeOpenGL_AutoInit_Construct()’未定义的引用
collect2: error: ld returned 1 exit status
CMakeFiles/camvox.dir/build.make:523: recipe for target '../camvox' failed
make[2]: *** [../camvox] Error 1
CMakeFiles/Makefile2:72: recipe for target 'CMakeFiles/camvox.dir/all' failed
make[1]: *** [CMakeFiles/camvox.dir/all] Error 2
Makefile:129: recipe for target 'all' failed
make: *** [all] Error 2

请问，你们给相机配的什么型号的镜头？

作者您好， 请问你们给这款相机配的什么镜头， 能否给个链接看看呢？

Hi there, I'm pretty appreciate your work, but I have trouble on evaluating with the groudtruth

I've downloaded your groundtruth.bag and convert the nav_msgs to poststamps in order to use evo to do the evaluation.
However, the issue happens and the groundtruth seems not like pictures in your paper, so would you please upload your groundtruth.txt.

catkin_make error

/home/chen/catkin_ws/src/CamVox/isee-preprocessing/src/isee_synchronize.cpp:39:8: error: ‘isee_synchronize’ was not declared in this scope
vector<isee_synchronize::GPS> rtk_datas;
^
/home/chen/catkin_ws/src/CamVox/isee-preprocessing/src/isee_synchronize.cpp:39:29: error: template argument 1 is invalid
vector<isee_synchronize::GPS> rtk_datas;
^
/home/chen/catkin_ws/src/CamVox/isee-preprocessing/src/isee_synchronize.cpp:39:29: error: template argument 2 is invalid
/home/chen/catkin_ws/src/CamVox/isee-preprocessing/src/isee_synchronize.cpp:75:23: error: ‘to_time’ declared as an ‘inline’ variable
inline double to_time(isee_synchronize::GPS rtk_data) // sensor_msgs::NavSatFix
^
/home/chen/catkin_ws/src/CamVox/isee-preprocessing/src/isee_synchronize.cpp:75:23: error: ‘double to_time’ redeclared as different kind of symbol
/home/chen/catkin_ws/src/CamVox/isee-preprocessing/src/isee_synchronize.cpp:70:15: note: previous declaration ‘double to_time(nav_msgs::Odometry)’
inline double to_time(nav_msgs::Odometry imu_data)
^
/home/chen/catkin_ws/src/CamVox/isee-preprocessing/src/isee_synchronize.cpp:75:23: error: ‘isee_synchronize’ has not been declared
inline double to_time(isee_synchronize::GPS rtk_data) // sensor_msgs::NavSatFix
^
/home/chen/catkin_ws/src/CamVox/isee-preprocessing/src/isee_synchronize.cpp:81:19: error: ‘isee_synchronize’ does not name a type
void rtkCbk(const isee_synchronize::GPS::ConstPtr &msg)
^
/home/chen/catkin_ws/src/CamVox/isee-preprocessing/src/isee_synchronize.cpp:81:40: error: expected unqualified-id before ‘::’ token
void rtkCbk(const isee_synchronize::GPS::ConstPtr &msg)

question related to construction of the range image from lidar pointcloud

Hi, thank you for the project. I have a question related to the construction of range image from livox lidar pointcloud here.
As I see in the code base, the range image has been created using pcl's createFromPointCloudWithFixedSize.

https://github.com/ISEE-Technology/CamVox/blob/main/isee-preprocessing/src/isee_synchronize.cpp#L335-L344
however, the width, height, camera matrix has been hardcoded here. Can you please elaborate about the choice of these values? Is this actually the parameters of the camera used in the experiment?

error when ./build.sh

because of the wrong VTK version?

/usr/include/vtk-6.2/vtkObjectBase.h:165:30: error: field ‘ReferenceCount’ has incomplete type ‘vtkAtomicInt’
vtkAtomicInt ReferenceCount;
^
In file included from /usr/include/vtk-6.2/vtkObjectBase.h:44:0,
from /usr/include/vtk-6.2/vtkSmartPointerBase.h:27,
from /usr/include/vtk-6.2/vtkSmartPointer.h:23,
from /usr/include/pcl-1.7/pcl/visualization/point_cloud_geometry_handlers.h:48,
from /usr/include/pcl-1.7/pcl/visualization/point_cloud_handlers.h:41,
from /usr/include/pcl-1.7/pcl/visualization/common/actor_map.h:40,
from /usr/include/pcl-1.7/pcl/visualization/pcl_visualizer.h:48,
from /usr/include/pcl-1.7/pcl/visualization/cloud_viewer.h:39,
from /home/hk/ws_CamVox/src/CamVox/isee-camvox/src/pointcloudmapping.cc:7:
/usr/include/vtk-6.2/vtkAtomicInt.h:307:7: note: declaration of ‘class vtkAtomicInt’
class vtkAtomicInt: public detail::vtkAtomicIntImpl
^
In file included from /home/hk/ws_CamVox/src/CamVox/isee-camvox/src/pointcloudmapping.cc:7:0:
/usr/include/pcl-1.7/pcl/visualization/cloud_viewer.h:202:14: warning: ‘template class std::auto_ptr’ is deprecated [-Wdeprecated-declarations]
std::auto_ptr<CloudViewer_impl> impl_;
^
In file included from /usr/include/c++/5/memory:81:0,
from /usr/include/c++/5/thread:40,
from /home/hk/ws_CamVox/src/CamVox/isee-camvox/include/System.h:19,
from /home/hk/ws_CamVox/src/CamVox/isee-camvox/include/pointcloudmapping.h:4,
from /home/hk/ws_CamVox/src/CamVox/isee-camvox/src/pointcloudmapping.cc:1:
/usr/include/c++/5/bits/unique_ptr.h:49:28: note: declared here
template class auto_ptr;
^
/home/hk/ws_CamVox/src/CamVox/isee-camvox/src/pointcloudmapping.cc: In member function ‘void Camvox::PointCloudMapping::updatecloud()’:
/home/hk/ws_CamVox/src/CamVox/isee-camvox/src/pointcloudmapping.cc:168:31: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]
for (int i = 0; i < currentvpKFs.size(); i++)
^
/home/hk/ws_CamVox/src/CamVox/isee-camvox/src/pointcloudmapping.cc:170:35: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]
for (int j = 0; j < pointcloud.size(); j++)
^
/home/hk/ws_CamVox/src/CamVox/isee-camvox/src/pointcloudmapping.cc:172:44: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]
if (pointcloud[j].pcID == currentvpKFs[i]->mnFrameId)
^
CMakeFiles/Camvox.dir/build.make:315: recipe for target 'CMakeFiles/Camvox.dir/src/pointcloudmapping.cc.o' failed
make[2]: *** [CMakeFiles/Camvox.dir/src/pointcloudmapping.cc.o] Error 1
CMakeFiles/Makefile2:98: recipe for target 'CMakeFiles/Camvox.dir/all' failed
make[1]: *** [CMakeFiles/Camvox.dir/all] Error 2
Makefile:102: recipe for target 'all' failed
make: *** [all] Error 2
Converting vocabulary to binary
./tools/bin_vocabulary: error while loading shared libraries: libDBoW2.so: cannot open shared object file: No such file or directory

Cannot access the bag files

Hi!
Hope you are well.
Thanks for sharing the code!
I am facing issues while accessing the bag files from zywok. Would it be possible to share the bag files using Google Drive?

TIA
Harshal

Can this work be used with other lidars and cameras?

Hello, thanks for your excellent work. I would like to test this work on my own sensors, such as the RoboSense lidar and ZED camera, is this feasible? If possible, how do I modify it?

Is it possible to achieve 10-5cm accuracy with this technology?

What would have to be done (improve IMU, Camera, GPS, algorithm, etc.) to achieve survey grade results with a similar setup?

can not open dataset download link

Why there would be a factor and an offset when computing range?

Dataset problem.

Thank you for your excellent work.
I downloaded the data set CamVox.bag you uploaded, but when I execute the following command:
rosbag play CamVox.bag
Display:
[ INFO] [1622026026.629494531]: Opening CamVox.bag [FATAL] [1622026026.629722490]: Required 'op' field missing
Then I execute the following command:
rosbag info CamVox.bag
Display:
ERROR bag unindexed: CamVox.bag. Run rosbag reindex.
Then I execute the command:
rosbag reindex CamVox.bag
After reindexing rosbag, I execute the command again:
rosbag info CamVox.bag
Display:
path: CamVox.bag version: 2.0 duration: 6:11s (371s) start: May 15 2020 16:14:29.30 (1589530469.30) end: May 15 2020 16:20:41.00 (1589530841.00) size: 15.0 GB messages: 7436 compression: none [7436/7436 chunks] types: sensor_msgs/Image [060021388200f6f0f447d0fcd9c64743] topics: /isee_depth 3718 msgs : sensor_msgs/Image /isee_rgb 3718 msgs : sensor_msgs/Image
As shown above, I lost /IMU and /livox_ lidar.
Can you upload the intact rosbag data again?
Thank You!

The problem Ubuntu18.04 compiling

I have encountered problems compiled in Ubuntu18.04,it seems to need PCL 1.7 for the project,but PCL 1.8 in ros melodic,how can i do？Thank you for your help!

关于内外参问题

您好，非常感谢你们的分享，这个工作非常有意义。
我仔细阅读了你们的文章和代码，我有一点关于内外参的问题想请教。
1 外参标定优化工作：在Calibrating::Run()中优化了best_r_, best_p_, best_y_，优化旋转是为了矫正生成深度图像的偏差，但我好像没发现使用其对激光生成的深度图像进行矫正，是不是我理解的有问题呢？
2 激光生成深度图像的工作：
1）在livox_lidar2depth函数中
这些变量“float fx=1732.78774,fy=1724.88401,cx=798.426021,cy=312.570668; ”是表示相机的内参么，好像与camera.yaml文件中
Camera.fx: 957.994
Camera.fy: 955.3280
Camera.cx: 790.335
Camera.cy: 250.6631 的参数不太一致，如果不是相机内参的话是什么意思呢。
2） 这个变量“Eigen::Affine3f sensorPose = (Eigen::Affine3f)Eigen::Translation3f(-0.082984,-0.0221759,0.0727962);”是坐标中心到相机的平移量么，平移对生成深度图影响较小所以在“Calibrating::Run()”中仅优化了旋转么？

期待您的回复，谢谢啦～ 再次感谢你们的分享

ground truth

请问可以公布ground truth数据吗