跑vins初始化时出现misalign visual structure with IMU很可能是因为订阅的IMU话题里加速度的值不对或者偏差太大

于 2024-07-12 20:56:23 发布
阅读量309 收藏 2
点赞数 10
分类专栏： 无人机-SLAM-vins 文章标签： vins
版权声明：本文为博主原创文章，遵循 CC 4.0 BY-SA 版权协议，转载请附上原文出处链接和本声明。
本文链接：https://blog.csdn.net/sinat_16643223/article/details/140388334
版权

跑vins初始化时出现misalign visual structure with IMU很可能是因为订阅的IMU话题里加速度的值不对或者偏差太大

这种可能出现在比如，用港大的MARS数据集跑vins，如果订阅Livox的imu话题/livox/imu，vins就会终端显示那样，初始化不过去，因为/livox/imu里的加速度值是普通加速度值的十分之一。。。可以改订阅MARS数据集里大疆飞控的imu话题。 

[ INFO] [1720780738.414808862]: misalign visual structure with IMU
[ INFO] [1720780738.549597874]: IMU excitation not enouth!
[ WARN] [1720780738.590833166]: gyroscope bias initial calibration -3.75238e-05     0.000124  4.95129e-05
[ INFO] [1720780738.602172607]: misalign visual structure with IMU
[ INFO] [1720780738.631519101]: IMU excitation not enouth!
[ WARN] [1720780738.698023176]: gyroscope bias initial calibration  -1.4687e-05  0.000237636 -1.50438e-05
[ INFO] [1720780738.709986578]: misalign visual structure with IMU
[ INFO] [1720780738.714009980]: IMU excitation not enouth!
[ INFO] [1720780738.714271771]: Not enough features or parallax; Move device around
[ INFO] [1720780738.801111269]: IMU excitation not enouth!
[ WARN] [1720780738.897142993]: gyroscope bias initial calibration -0.000655604   0.00639225 -0.000715006
[ INFO] [1720780738.908234825]: misalign visual structure with IMU
[ INFO] [1720780738.911420583]: IMU excitation not enouth!
[ WARN] [1720780739.022555211]: gyroscope bias initial calibration -0.000759697 -0.000721095  -0.00116168
[ INFO] [1720780739.034211427]: misalign visual structure with IMU
[ INFO] [1720780739.214047618]: IMU excitation not enouth!
[ WARN] [1720780739.263015890]: gyroscope bias initial calibration  0.00287479 -0.00853525 0.000172261
[ INFO] [1720780739.279455011]: misalign visual structure with IMU
[ INFO] [1720780739.531084662]: IMU excitation not enouth!
[ WARN] [1720780739.603870875]: gyroscope bias initial calibration -7.61861e-05   0.00512783    0.0060488
[ INFO] [1720780739.613776579]: misalign visual structure with IMU
[ INFO] [1720780739.852045912]: IMU excitation not enouth!
[ INFO] [1720780740.185337109]: IMU excitation not enouth!
[ INFO] [1720780740.513487602]: IMU excitation not enouth!
[ INFO] [1720780740.542914847]: Not enough features or parallax; Move device around
[ INFO] [1720780740.823974256]: IMU excitation not enouth!
W0712 18:39:00.898424 15362 levenberg_marquardt_strategy.cc:115] Linear solver failure. Failed to compute a step: Eigen failure. Unable to perform dense Cholesky factorization.
W0712 18:39:00.900166 15362 levenberg_marquardt_strategy.cc:115] Linear solver failure. Failed to compute a step: Eigen failure. Unable to perform dense Cholesky factorization.
W0712 18:39:00.900640 15362 levenberg_marquardt_strategy.cc:115] Linear solver failure. Failed to compute a step: Eigen failure. Unable to perform dense Cholesky factorization.
W0712 18:39:00.902647 15362 levenberg_marquardt_strategy.cc:115] Linear solver failure. Failed to compute a step: Eigen failure. Unable to perform dense Cholesky factorization.
W0712 18:39:00.904049 15362 levenberg_marquardt_strategy.cc:115] Linear solver failure. Failed to compute a step: Eigen failure. Unable to perform dense Cholesky factorization.
W0712 18:39:00.905723 15362 levenberg_marquardt_strategy.cc:115] Linear solver failure. Failed to compute a step: Eigen failure. Unable to perform dense Cholesky factorization.
W0712 18:39:00.906128 15362 levenberg_marquardt_strategy.cc:115] Linear solver failure. Failed to compute a step: Eigen failure. Unable to perform dense Cholesky factorization.
W0712 18:39:00.908887 15362 levenberg_marquardt_strategy.cc:115] Linear solver failure. Failed to compute a step: Eigen failure. Unable to perform dense Cholesky factorization.
W0712 18:39:00.909301 15362 levenberg_marquardt_strategy.cc:115] Linear solver failure. Failed to compute a step: Eigen failure. Unable to perform dense Cholesky factorization.
W0712 18:39:00.911664 15362 levenberg_marquardt_strategy.cc:115] Linear solver failure. Failed to compute a step: Eigen failure. Unable to perform dense Cholesky factorization.
[ INFO] [1720780741.146508554]: IMU excitation not enouth!
[ WARN] [1720780741.226884079]: gyroscope bias initial calibration  -0.0022124  0.00389549 -0.00745966
[ INFO] [1720780741.232613543]: misalign visual structure with IMU
[ INFO] [1720780741.487772556]: IMU excitation not enouth!
W0712 18:39:01.564990 15362 levenberg_marquardt_strategy.cc:115] Linear solver failure. Failed to compute a step: Eigen failure. Unable to perform dense Cholesky factorization.
W0712 18:39:01.569908 15362 levenberg_marquardt_strategy.cc:115] Linear solver failure. Failed to compute a step: Eigen failure. Unable to perform dense Cholesky factorization.
W0712 18:39:01.571506 15362 levenberg_marquardt_strategy.cc:115] Linear solver failure. Failed to compute a step: Eigen failure. Unable to perform dense Cholesky factorization.
[ INFO] [1720780741.815859911]: IMU excitation not enouth!
[ WARN] [1720780741.896610849]: gyroscope bias initial calibration 0.00324116 -0.0463724 0.00212543
[ INFO] [1720780741.900815278]: misalign visual structure with IMU
[ INFO] [1720780742.022928917]: IMU excitation not enouth!
[ INFO] [1720780742.049445325]: Not enough features or parallax; Move device around
[ INFO] [1720780742.219433378]: IMU excitation not enouth!
[ INFO] [1720780742.242262606]: Not enough features or parallax; Move device around
[ INFO] [1720780742.309897692]: IMU excitation not enouth!
[ INFO] [1720780742.327884906]: Not enough features or parallax; Move device around
[ INFO] [1720780742.395098470]: IMU excitation not enouth!
[ INFO] [1720780742.412748337]: Not enough features or parallax; Move device around
[ INFO] [1720780742.483289826]: IMU excitation not enouth!
[ INFO] [1720780742.500288117]: Not enough features or parallax; Move device around
[ INFO] [1720780742.689883587]: IMU excitation not enouth!
W0712 18:39:02.779299 15362 levenberg_marquardt_strategy.cc:115] Linear solver failure. Failed to compute a step: Eigen failure. Unable to perform dense Cholesky factorization.
[ INFO] [1720780742.896491149]: IMU excitation not enouth!
[ INFO] [1720780742.915764217]: Not enough features or parallax; Move device around
[ INFO] [1720780743.096273792]: IMU excitation not enouth!
[ INFO] [1720780743.106601449]: Not enough features or parallax; Move device around
[ INFO] [1720780743.173798626]: IMU excitation not enouth!
[ INFO] [1720780743.181246128]: Not enough features or parallax; Move device around
[ INFO] [1720780743.490377966]: IMU excitation not enouth!
[ INFO] [1720780743.495506295]: Not enough features or parallax; Move device around
[ INFO] [1720780743.808472104]: IMU excitation not enouth!
[ INFO] [1720780743.814476072]: Not enough features or parallax; Move device around
[ INFO] [1720780744.133004172]: IMU excitation not enouth!
[ INFO] [1720780744.134063923]: Not enough features or parallax; Move device around
[ INFO] [1720780744.443132535]: IMU excitation not enouth!
[ INFO] [1720780744.445363737]: Not enough features or parallax; Move device around
[ INFO] [1720780744.747217623]: IMU excitation not enouth!
[ INFO] [1720780744.749076143]: Not enough features or parallax; Move device around
[ INFO] [1720780745.083726229]: IMU excitation not enouth!
[ INFO] [1720780745.086242826]: Not enough features or parallax; Move device around
[ INFO] [1720780745.281053009]: IMU excitation not enouth!
[ INFO] [1720780745.282869760]: Not enough features or parallax; Move device around
[ INFO] [1720780745.606142630]: IMU excitation not enouth!
[ INFO] [1720780745.611400070]: Not enough features or parallax; Move device around
[ INFO] [1720780745.914894021]: IMU excitation not enouth!
[ INFO] [1720780745.923654104]: Not enough features or parallax; Move device around
[ INFO] [1720780746.225399174]: IMU excitation not enouth!
[ INFO] [1720780746.238673994]: Not enough features or parallax; Move device around
[ INFO] [1720780746.534827952]: IMU excitation not enouth!
[ INFO] [1720780746.562708694]: Not enough features or parallax; Move device around
[ INFO] [1720780746.848697223]: IMU excitation not enouth!
unstable features tracking, please slowly move you device!
W0712 18:39:06.921356 15362 levenberg_marquardt_strategy.cc:115] Linear solver failure. Failed to compute a step: Eigen failure. Unable to perform dense Cholesky factorization.
W0712 18:39:06.922619 15362 levenberg_marquardt_strategy.cc:115] Linear solver failure. Failed to compute a step: Eigen failure. Unable to perform dense Cholesky factorization.
W0712 18:39:06.924743 15362 levenberg_marquardt_strategy.cc:115] Linear solver failure. Failed to compute a step: Eigen failure. Unable to perform dense Cholesky factorization.
W0712 18:39:06.925096 15362 levenberg_marquardt_strategy.cc:115] Linear solver failure. Failed to compute a step: Eigen failure. Unable to perform dense Cholesky factorization.
[ INFO] [1720780747.151864762]: IMU excitation not enouth!
W0712 18:39:07.213977 15362 levenberg_marquardt_strategy.cc:115] Linear solver failure. Failed to compute a step: Eigen failure. Unable to perform dense Cholesky factorization.
[ INFO] [1720780747.459902332]: IMU excitation not enouth!
[ INFO] [1720780747.479960354]: Not enough features or parallax; Move device around
[ INFO] [1720780747.759867956]: IMU excitation not enouth!
[ INFO] [1720780747.778443854]: Not enough features or parallax; Move device around
[ INFO] [1720780748.078044134]: IMU excitation not enouth!
[ INFO] [1720780748.098868041]: Not enough features or parallax; Move device around
[ INFO] [1720780748.382774742]: IMU excitation not enouth!
[ INFO] [1720780748.411633912]: Not enough features or parallax; Move device around
[ INFO] [1720780748.687422224]: IMU excitation not enouth!
[ INFO] [1720780748.715232734]: Not enough features or parallax; Move device around
[ INFO] [1720780748.879523789]: IMU excitation not enouth!
[ INFO] [1720780748.910619281]: Not enough features or parallax; Move device around
[ INFO] [1720780749.065306618]: IMU excitation not enouth!
[ INFO] [1720780749.093664152]: Not enough features or parallax; Move device around
[ INFO] [1720780749.144901805]: IMU excitation not enouth!
[ INFO] [1720780749.172483439]: Not enough features or parallax; Move device around
[ INFO] [1720780749.233487051]: IMU excitation not enouth!
[ INFO] [1720780749.258838074]: Not enough features or parallax; Move device around
[ INFO] [1720780749.318450241]: IMU excitation not enouth!
[ INFO] [1720780749.333940902]: Not enough features or parallax; Move device around
[ INFO] [1720780749.516716499]: IMU excitation not enouth!
[ INFO] [1720780749.533863910]: Not enough features or parallax; Move device around
[ INFO] [1720780749.817373080]: IMU excitation not enouth!
[ INFO] [1720780749.828832999]: Not enough features or parallax; Move device around
[ INFO] [1720780750.128388739]: IMU excitation not enouth!
[ INFO] [1720780750.448110860]: IMU excitation not enouth!
[ INFO] [1720780750.465091684]: Not enough features or parallax; Move device around
[ INFO] [1720780750.759667079]: IMU excitation not enouth!
[ INFO] [1720780750.772518017]: Not enough features or parallax; Move device around
[ INFO] [1720780751.074353066]: IMU excitation not enouth!
[ INFO] [1720780751.079510439]: Not enough features or parallax; Move device around
[ INFO] [1720780751.373883311]: IMU excitation not enouth!
[ INFO] [1720780751.389610160]: Not enough features or parallax; Move device around
[ INFO] [1720780751.684095434]: IMU excitation not enouth!
[ INFO] [1720780751.695431188]: Not enough features or parallax; Move device around

也是受到这个人回答的启发，这个人看了下也是港科大的 

https://github.com/HKUST-Aerial-Robotics/VINS-Mono/issues/333

https://github.com/hkust-aerial-robotics/vins-mono/issues/62

 

 

诗筱涵
关注
专栏目录
VINS-Mono代码学习记录（八）--processImage( )续(包含初始化）
weixin_38504276的博客
11-27 578
上一节其实已经学到了初始化的部分，就是求解相机和IMU的外参数：旋转。 接下来继续吧… 在processImage( )中，涉及到初始化的代码为： //[4]判断是初始化还是非线性优化 if (solver_flag == INITIAL)//初始化 { if (frame_count == WINDOW_SIZE)//如果帧数已经到达滑动窗口设定的帧数，就进行优...
vins中imu融合_VINS代码解读
weixin_28871743的博客
12-30 928
VINS_estimator摘抄我们初始化的原因是单目惯性紧耦合系统是一个非线性程度很高的系统，首先单目是无法获得空间中的绝对尺度，而IMU又必然存在偏置，在后面进行求解的时候还需要用到重力加速度(包括大小和方向)，对于速度比较敏感的条件下，比如说无人机，又要精确的速度信息，因此，如何有效的在紧耦合系统处理之前计算出这些量，对整个紧耦合系统的鲁棒性有着重大的意义(其实这里就可以理解成相机标定一样，...
VINS 估计器之结构初始化
weixin_34347651的博客
04-19 382
为什么要初始化 非线性VINS估计器的性能对于初始的速度，尺度，重力向量，空间点3D位置，以及外参等非常敏感。在很多场合中，能做到相机和IMU即插即用，线上自动校准与初始化，将会给用户带来极大的方便性。VINS里面分四步进行，第一个就是上次讲的旋转外参校准，第二个就是找到某帧作为系统初始化原点，计算3D地图点，第三就是将相机坐标系转到IMU坐标系中，第四就是相机与IMU对齐，包括IMU零偏初始化，...
mono touch ：Displaying Alerts
每天进步一点点
12-14 1163
Displaying Alerts A common user interface task is displaying an alert, either to advise the user of an error condition or make a choice. The simplest alert contains a single button, as shown in thi
ceres之cholesky求解器
xiechaoyi123的博客
06-15 1665
参考代码： normal_sparse_cholesky:https://github.com/ceres-solver/ceres-solver/blob/master/internal/ceres/dynamic_sparse_normal_cholesky_solver.cc dense_cholesky:https://github.com/ceres-solver/ceres-solver/blob/master/internal/ceres/dense_normal_cholesky_sol
ceres之LM算法
xiechaoyi123的博客
06-20 4881
Ceres作为一个优化算法库，在许多领域中有着至关重要的作用，比如slam系统中的优化问题-集束调整BA，就可以通过Ceres去实现，官方文档地址：http://ceres-solver.org/nnls_tutorial.html#bundle-adjustment 本文主要是解析ceres中的LM算法过程，参考代码地址： https://github.com/ceres-solver/ceres-solver/tree/master/internal/ceres 一、主要流程 先贴个图，L.
视觉SLAM十四讲CH9代码解析及课后习题详解
生活如此美好，致敬每一个热爱“热爱”的人！
11-22 6493
点云查看工具CloudCompare安装: 方法1： 命令行安装： sudo snap install cloudcompare 安装完成后，可以将 ccViewer和CloudCompare添加到任务栏里，也可以使用下面的命令行打开： 打开ccViewer： cloudcompare.ccViewer 打开CloudCompare： cloudcompare.CloudCompare 方法2： 在软件中心中进行安装： ...
VINS-Mono之外参标定和视觉IMU联合初始化
Hansry的博客
02-18 2271
文章目录1.前言2. 利用旋转约束估计外参数旋转qcbq^{b}_{c}qcb​ 1.前言 本博客主要介绍VINS-Mono初始化时相机与IMU对齐，主要包括相机到IMU的外参估计、陀螺仪偏置、相机位移和估计空间点的尺度、重力加速度、每帧速度，主要参考了深蓝学院的VIO课程及博客VINS-Mono理论学习——视觉惯性联合初始化与外参标定和崔华坤的《VINS 论文推导及代码解析》。 总体而言，视觉I...
vins中imu融合_VINS的初始化与camera-imu外参标定
weixin_39669147的博客
12-21 456
1. 外参旋转矩阵初始化在VINS的Feature Detection and Tracking模块中，用Harris特征匹配，通过F矩阵和RANSAC算法恢复出 ；相应地使用IMU陀螺仪数据预积分得到 ;这两个测量量满足： * = * 将测量得到的R用四元数表示，并用四元数的运算(左乘右乘)得到约束方程： * = 0用多组数据得到超定方程： * = 0通过求最小二乘解，得到 ， ...
imu相机标定_VINS（四）初始化与相机IMU外参标定
weixin_29498577的博客
12-29 670
和单目纯视觉的初始化只需要获取R，t和feature的深度不同，VIO的初始化话通常需要标定出所有的关键参数，包括速度，重力方向，feature深度，以及相机IMU外参$R_{c}^{b}$和$p_{c}^{b}$。一. 外参旋转矩阵初始化在Feature Detection and Tracking模块中，利用Harris特征点匹配通过基础矩阵和Ransac恢复出$R_{Ck+1}^{Ck}$；...
VINS 中的IMU预积分推导和代码解读.zip
06-06
在自动驾驶领域，视觉惯性导航系统（Visual-Inertial Navigation System，简称VINS）是一种重要的定位技术，它结合了摄像头的视觉信息与惯性测量单元（IMU）的数据，为移动平台提供高精度的实时定位。本文将深入探讨...
VINS ---初始化
zwlq1314521的专栏
07-21 1080
通过匹配的视觉特征点 对极约束获取本质矩阵E，再分解本质矩阵获得相机的R，T bool MotionEstimator::solveRelativeRT(const vector> &corres, Matrix3d &Rotation, Vector3d &Translation) { if (corres.size() >= 15) { vector ll,
SLAM学习（三）——三维空间刚体运动（实践）
qq_44787464的博客
08-27 394
代码库：http://github.com/gaoxiang12 1.实践：Eigen 上篇文章主要写了三维刚体运动的一些理论知识。 这部分实践主要介绍使用Eigen来表示矩阵、向量，随后引申至旋转矩阵和变换矩阵的计算。 在实践之前，您需要安装Eigen库，网上有很多安装教程，相信难不倒大家，然后就废话少说，直接上代码： 代码文件名：eigenMatrix.cpp #include <ios...
Eigen解线性方程组
weixin_38009585的博客
09-03 4391
记录一下 一. 矩阵分解： 矩阵分解 (decomposition, factorization)是将矩阵拆解为数个矩阵的乘积，可分为三角分解、满秩分解、QR分解、Jordan分解和SVD（奇异值）分解等，常见的有三种：1)三角分解法 (Triangular Factorization)，2)QR 分解法 (QR Factorization)，3)奇异值分解法 (Singular Value ...
Eigen: 遇到的报错或中断及解决方法
qq_37896521的博客
09-11 2038
是是是
VINS初始化与特征点跟踪在Linux上的OracleHA双机热备实践
初始化阶段，VINS采用视觉和IMU的松耦合方案，通过Structure From Motion (SFM)方法求解所有帧的位姿和路标点的3D位置，再与IMU预积分值对齐，以此估计重力方向、尺度因子、陀螺仪bias以及每帧的速度。 "IMU预积分...
评论
添加红包

请填写红包祝福语或标题

个

红包个数最小为10个

元

红包金额最低5元

当前余额3.43元 前往充值 >
需支付：10.00元
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付元
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明：

1.余额是钱包充值的虚拟货币，按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载，可以购买VIP、付费专栏及课程。

余额充值