alexgecontrol / sensor-fusion-for-localization-courseware Goto Github PK

本地ubuntu部署环境,按照cpu.Dockerfile中指导的版本进行安装,编译成功,但是除了master能运行成功,其他分支均报内存异常.
process[data_pretreat_node-2]: started with pid [31362]
process[scan_scan_registration_node-3]: started with pid [31363]
process[scan_map_registration_node-4]: started with pid [31364]
process[evaluation_node-5]: started with pid [31365]
process[rviz-6]: started with pid [31366]
terminate called after throwing an instance of 'std::bad_alloc'
what(): std::bad_alloc
terminate called after throwing an instance of 'std::bad_alloc'
what(): std::bad_alloc
terminate called after throwing an instance of 'terminate called after throwing an instance of 'std::bad_alloc'

具体描述

第二章直接法推导最优位姿时使用的右扰动模型，但是第三章又使用了左扰动模型，请问哪一种是适合在帧间匹配时使用？

第四章代码框架在使用时，每次一到g2o优化地图的地方，就会报segmentation fault的错误。比如调用service 优化地图，或者不调用service，但是bag跑到300多秒的时候，由于会自动调用g2o optimize,都会出现segmentation fault。

第一章中网页的虚拟机屏幕分辨率低
文档说配置参数，查一圈还是不知道怎么配置，
想让这个分辨率改一下

Describe the bug
关于雷达运动补偿的公式图示理解

我画了个图，黑色线代表T0 Ti Pi在世界系下的含义，红色线代表对上图红色框里公式的理解

请问T0 Ti Pi含义的理解是否正确？公式推导的理解是否正确？

Error response from daemon: Get "https://registry.cn-shanghai.aliyuncs.com/v2/": unauthorized: authentication required

编程作业
能不能顺便把所有章节第一期的作业都放上来，想多花时间挑战一下

Screenshots
If applicable, add screenshots to help explain your problem.

通过clone新的github上面的工程，再新的下载根目录中查看运行的容器显示没有
itcast@ubuntu:/Sensor-Fusion-for-Localization-Courseware$ docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
执行docker-compose down && docker-compose up报错如下
itcast@ubuntu:/Sensor-Fusion-for-Localization-Courseware$ docker-compose down && docker-compose up
Removing network sensor-fusion-for-localization-courseware_default
Creating network "sensor-fusion-for-localization-courseware_default" with the default driver
Creating sensor-fusion-workspace-cpu ... error

ERROR: for sensor-fusion-workspace-cpu Cannot create container for service workspace-bionic-cpu-vnc: Conflict. The container name "/sensor-fusion-workspace-cpu" is already in use by container "934ce78a200990466c077e2c186a3f39b7a8e275d9caed8f0149cc80d920cc4f". You have to remove (or rename) that container to be able to reuse that name.

ERROR: for workspace-bionic-cpu-vnc Cannot create container for service workspace-bionic-cpu-vnc: Conflict. The container name "/sensor-fusion-workspace-cpu" is already in use by container "934ce78a200990466c077e2c186a3f39b7a8e275d9caed8f0149cc80d920cc4f". You have to remove (or rename) that container to be able to reuse that name.
ERROR: Encountered errors while bringing up the project.
请问老师这个怎么处理？

文档说配置参数，查一圈还是不知道怎么配置，想让这个分辨率改一下

老师好，我把#include "lidar_localization/models/registration/sicp/ICP.h" 从 scip_registration.cpp移动到 scip_registration.hpp下，编译时产生多次定义的错误 ，需要注释掉ICP.h中的 IWICP 的命名空间才能通过，是只有我电脑有这个问题吗（虽然并不影响结果），谢谢老师。

老师，第四章代码未作修改编译后，roslaunch lidar_localization mapping.launch 运行到一半后端节点报错，且transformed_odom不再更新，gnss轨迹能更新，如图所示，未做修改出现该问题原因是和电脑有关嘛？用的本地环境，装的库和readme中版本一致。谢谢老师解答。

Describe the bug
给repo名称前加了一个01序号，然后原来的docker就无法启动了，报了如下的错误，请问如何解决？

Screenshots

老师，请问自己实现的icp_svd轨迹前面还正常，后面尺度就越飘越大，有什么解决办法么？