《视觉SLAM14讲》第七讲pose_estimation_3d2d程序运行报错
在运行《视觉SLAM14讲》第七讲程序的时候,出现了报错,说
In file included from /usr/local/include/g2o/core/base_fixed_sized_edge.h:278:0,
from /usr/local/include/g2o/core/base_unary_edge.h:30,
from /home/zyf/zyf/slam14/slambook_zyf/ch7/pose_estimation_3d2d.cpp:8:
/usr/local/include/g2o/core/base_fixed_sized_edge.hpp: In instantiation of ‘void g2o::BaseFixedSizedEdge<D, E, VertexTypes>::linearizeOplusN() [with int N = 0; int D = 2; E = Eigen::Matrix<double, 2, 1>; VertexTypes = {VertexPose}]’:
/usr/local/include/g2o/core/base_fixed_sized_edge.hpp:191:41: required from ‘void g2o::BaseFixedSizedEdge<D, E, VertexTypes>::linearizeOplusNs(std::index_sequence<Ints ...>) [with long unsigned int ...Ints = {0}; int D = 2; E = Eigen::Matrix<double, 2, 1>; VertexTypes = {VertexPose}; std::index_sequence<Ints ...> = std::integer_sequence<long unsigned int, 0>]’
/usr/local/include/g2o/core/base_fixed_sized_edge.hpp:199:19: required from ‘void g2o::BaseFixedSizedEdge<D, E, VertexTypes>::linearizeOplus() [with int D = 2; E = Eigen::Matrix<double, 2, 1>; VertexTypes = {VertexPose}]’
/home/zyf/zyf/slam14/slambook_zyf/ch7/pose_estimation_3d2d.cpp:359:1: required from here
/usr/local/include/g2o/core/base_fixed_sized_edge.hpp:165:14: error: ‘FixedArray {aka class ceres::internal::FixedArray<double, 6>}’ has no member named ‘fill’
add_vertex.fill(0.);
~~~~~~~~~~~^~~~
/usr/local/include/g2o/core/base_fixed_sized_edge.hpp:172:30: error: ‘FixedArray {aka class ceres::internal::FixedArray<double, 6>}’ has no member named ‘data’
vertex->oplus(add_vertex.data());
~~~~~~~~~~~^~~~
/usr/local/include/g2o/core/base_fixed_sized_edge.hpp:178:30: error: ‘FixedArray {aka class ceres::internal::FixedArray<double, 6>}’ has no member named ‘data’
vertex->oplus(add_vertex.data());
~~~~~~~~~~~^~~~
我也不清楚是怎么回事,重新安装了一遍g2o的库也不行,大佬们给看看是怎么回事
...全文
相关推荐
kitti数据pose
at Chicago的缩写,是计算机视觉领域中一个广泛使用的开源数据集,尤其在自动驾驶、3D物体检测、视觉定位(pose estimation)等领域具有重要价值。
Desktop_avalish_
. **3D Pose Estimation**:在三维空间中估计物体的位置和旋转,是机器人导航和自动驾驶的关键部分。5.
3D_LIDAR PAPER+ imu.zip
3D_LIDAR PAPER+ imu.zip 这个压缩包包含了一系列关于3D激光雷达(LIDAR)SLAM(Simultaneous Localization And Mapping,即同时定位与建图
slam-python:SLAM-使用OpenCV和NumPy同时进行本地化和映射
在这个过程中,我们通常会遇到两个主要问题:位姿估计(pose estimation)和地图构建(mapping)。OpenCV库可以处理图像处理和特征检测,这些是解决这两个问题的关键步骤。
Relative Pose Estimation Using Non-overlapping Multi-camera Clusters
这些方法不仅解决了传统SLAM技术在特定条件下无法应用的问题,还为未来的计算机视觉和机器人学研究提供了新的思路和技术支撑。


