clothooo / lvt2calib Goto Github PK
View Code? Open in Web Editor NEWA unified calibration for 3D LiDARs, visual cameras and thermal cameras
A unified calibration for 3D LiDARs, visual cameras and thermal cameras
您好
想请教个问题,我测试了livox avia+红外相机,参数设置没有问题。当我采用mid70时候,发现rviz无法显示4个圆点云,终端也显示无法检测到标定板子。
1. 根据您help里面介绍的方法,更改了i_filter_out_max参数5 10 20 30 40...100不起作用。
2. 将cluster_size_min从0.05改为0.04点云出现了,等了1,2分钟开始收集,但是激光雷达数据收集终端到60帧不停止,一直在收集,这个调参数有什么需要注意的,Recommend cluster_size_max as 2~5 and match the value of cluster_tole.看还需要兼顾cluster_tole,看代码里面cluster_size_max设置的是20。
3. 刚准备向您请教,然后我在突然改了上述参数,马上开始收集就可以到60帧停止了。
4. 还有上次关于4圆孔点云模板的问题,我自己用cloudcompare剪裁了一个模板,您这个模板有坐标系要求吗
Found this error when I updated the code. I think you might have forgot to push the src/lidar/robosense_pattern.cpp on branch ros_noetic.
CMake Error at lvt2calib/CMakeLists.txt:257 (add_executable):
Cannot find source file:
src/lidar/robosense_pattern.cpp
Tried extensions .c .C .c++ .cc .cpp .cxx .cu .m .M .mm .h .hh .h++ .hm
.hpp .hxx .in .txx
CMake Error at lvt2calib/CMakeLists.txt:257 (add_executable):
No SOURCES given to target: robosense_pattern
CMake Generate step failed. Build files cannot be regenerated correctly.
Invoking "cmake" failed
Hello, im running your code on the docker container provided and it doesn't find the calibration board. I've already tried changing the parameters with rqt_reconfigure but nothing happens anyways.
I have even filtered the data to cutout everything that is not the calib board but it doesnt work.
Can you help me looking if something on my data is wrong or if i'm doing something wrong as well?
Here is the link for my rosbag with my lidar data:
https://drive.google.com/file/d/1OFpjm0hbMcirOwzK1v-oFvXuiEsHwzCb/view?usp=sharing
I also uploaded the raw data, without filtering the points.
https://drive.google.com/file/d/1few2nxKRxKD6GLpGDpSSr7B6tggUZp18/view?usp=sharing
Thank you for providing such an excellent project for multi-sensor calibration!
However, I only have Hesai lidars at the moment so I can not try it out on real sensors. Do you have any suggestions on porting it to Hesai lidars? For example, could you note the difference between the three lidar brands?
Sorry if this is a dumb question, but how can I give my data as input for your LiDAR-LiDAR calibration code? I'm running your code using docker, with the "pull_run.sh" script and have a os1-32 sensor.
Would this work with the Livox MID360 Lidar, too? What settings would have to be changed?
Thanks!
I am in a Building where everything is black and white and the program is not able to find the calib board on the image. So best option for me would be red. Is the program able to detect a red board?
您好,请问在您的工作中两个相机的内参是如何标定的,是用您2018年在ICARCV中two step论文中提到的方法吗?以及如果用张正友平面标定法进行热成像和可见光相机的双目标定时图像的分辨率不同该如何处理?
Hello,
after 2 or 3 positions the software prints MAX_FRAME REACHED and simply stops. There is no error or warning message. Unfortunately is the result after 2 or 3 positions not precise enough. Is this a bug or am I doing something wrong here.
I am using the ubuntu 20.04/noetic version.
您好,想请教您几个问题
1.咱们这个标定板长可以用1000毫米吗,其他尺寸不变,我这边担心大一点容易弯曲,我用铝合金板材也比较重。
2.咱们的标定算法适用激光雷达和红外相机的标定吧,我采用红色的标定板可以实现吗,我看之前有人有疑问。
期待您的回复!
I'm using ouster os1_32 for calibration.
board isn't recognized on the side of lidar, but front and back working.
How can I fix it?
请问您的论文在哪里可以找到?
ICP has converged, score is 0.0618294
the mean is 0.156037
the mean is 0.249597
[ INFO] [1695296928.393437322]: The rmse of unknown cloud to template cloud is: 0.193602
[ INFO] [1695296928.393453946]: The rmse of template cloud to unknown cloud is: 0.147167
[ INFO] [1695296928.393462795]: The mean of rmse is: 0.170385
[ INFO] [1695296928.393539736]: Remianing 708 points in cloud
[ WARN] [1695296928.393831620]: Segmentation No.4 spend[0.000000ms]
number of point clouds in the plane: 410
min_to_rm: 0 max_to_rm: 30
size after intensity filter: 59
size of boudary: 57
[pcl::IterativeClosestPoint::computeTransformation] Not enough correspondences found. Relax your threshold parameters.
[ WARN] [1695296928.402370885]: ICP hasn't converged!
the mean is 1.7786
the mean is 1.67792
[ INFO] [1695296928.404290810]: The rmse of unknown cloud to template cloud is: 0.409431
[ INFO] [1695296928.404315916]: The rmse of template cloud to unknown cloud is: 0.314987
[ INFO] [1695296928.404328926]: The mean of rmse is: 0.362209
[ INFO] [1695296928.404423313]: Remianing 298 points in cloud
[ WARN] [1695296928.404860121]: PointCloud represneting the Cluster: 2303 data points.
[ WARN] [1695296928.409570158]: Segmentation No.5 spend[4.000000ms]
number of point clouds in the plane: 829
min_to_rm: 0 max_to_rm: 30
size after intensity filter: 764
size of boudary: 318
您好,请问,终端里一直输出上面的内容,在rviz里没有激光点云的标定版平面,终端里也一直没有velo_16的数字累积。我用的是16线的velodyne激光,标定版与您提到的一致。请问该如何解决?如何调整参数?
[os1_32/laser_pattern-1] process has died [pid 55937, exit code -11, cmd /home/chandroidx/catkin_ws/devel/lib/lvt2calib/ouster_pattern ~cloud_laser:=/ouster/points __name:=laser_pattern __log:=/home/chandroidx/.ros/log/f296362a-e67b-11ee-80c1-b3f7544c7ef2/os1_32-laser_pattern-1.log].
log file: /home/chandroidx/.ros/log/f296362a-e67b-11ee-80c1-b3f7544c7ef2/os1_32-laser_pattern-1*.log
it seems occurs on
// AutoDetectLaser.h
plane_segmentation.segment (*inliers, *coefficients);
您好,我在使用代码运行您的livox_rgb_thermal_rosbag数据集时,在雷达到相机的标定模块里,圆形特征点云在rviz中不能显示。按照您HELP.md中设置“i_filter_out_max”参数后,“/livox_horizon/laser_pattern/calib_board_cloud”能显示,但是“/(ns_lidar)/laser_pattern_circle/circle_center_cloud”始终显示不了,同时在T2终端里,一直显示“cannot find cailb borad”,请问在雷达提取的部分,参数需要怎么调试,能提供一下具体调试的方法吗?我在设置i_filter_out_max时,是在rviz中选取了标定板和支架强度的中间值。
您好!
又得麻烦一下,想请教个问题,看能帮忙看一下吗
在适配速腾32线激光雷达时,我利用rs_to_velodyne将速腾32线转为velodyne格式,直接用vel_32格式雷达运行代码,遇到检测不到点云模板
都是激光雷达和红外相机标定,其他终端都没有问题
1.我将转换后的雷达格式改为XYZIRT,在检测点云的终端遇到了下面报错,直接挂了
2.然后我将雷达格式改为XYZI,不出现上述报错,但是调整了i_filter_out_max、cluster_size_min、rmse_ukn2tpl_thre' and 'rmse_tpl2ukn_thre'都无法解决
3.上述两个错误都存在咱们这个项目需要匹配"ring"和"range"参数吗
雷达数据格式
见下图
下面是我测试用的三个数据,80fov.bag是激光雷达只有80度视角,XYZIRT.bag是XYZIRT格式的雷达数据,XYZI.bag是XYZI格式的雷达数据
链接: https://pan.baidu.com/s/1_Ouz7qn_lZRsvYBd3a1fpA?pwd=ui72 提取码: ui72
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.