全流程:ubuntu20.04运行Vins-fusion+ego_planner

求上进的小怪兽 2024-07-24 14:37:01 阅读 72

目录

一、下载编译Vins-fusion

           1. 创建工作空间vins_ws,并在src目录下git源码并编译

           2. 修改Cmakelist (四个文件夹)

           3. 添加头文件

           4. 仍然有一部分错误:

 二、编译通过!测试运行:

           1. 运行数据集Euroc

           2. 运行OAK-D 双目相机

 *补充:

三、VIns-Fusion配置参数矫正

 1. 启动Vins程序,查看输出话题

 2. 终端输入,查看程序输出position的漂移情况

 3. (将相机与板载计算机固定在无人机上之后)抱起无人机在场地内缓慢走动,在vins的输出文件夹下打开extrinsic_parameter.txt


首先,安装相关环境依赖ceres solver、eigen以及ros-noetic版本(教程很多,此处省略)。

注:ubuntu 20.04会自带一个opencv 4.2版本。

一、下载编译Vins-fusion

1. 创建工作空间vins_ws,并在src目录下git源码并编译:

git clone https://github.com/HKUST-Aerial-Robotics/VINS-Fusion.git

cd ..

catkin_make

此时,编译会出现一系列错误:

<code>make[2]: *** [VINS-Fusion/camera_models/CMakeFiles/Calibrations.dir/build.make:128:VINS-Fusion/camera_models/CMakeFiles/Calibrations.dir/src/camera_models/CostFunctionFactory.cc.o] 错误 1

make[1]: *** [CMakeFiles/Makefile2:843:VINS-Fusion/camera_models/CMakeFiles/Calibrations.dir/all] 错误 2

In file included from /usr/local/include/ceres/autodiff_cost_function.h:130,

from /usr/local/include/ceres/ceres.h:37,

from /home/fastpx4/dai_ws/src/VINS-Fusion/global_fusion/src/globalOpt.h:20,

from /home/fastpx4/dai_ws/src/VINS-Fusion/global_fusion/src/globalOpt.cpp:12:

/usr/local/include/ceres/internal/autodiff.h: In instantiation of ‘bool ceres::internal::AutoDifferentiate(const Functor&, const T* const*, int, T*, T**) [with int kNumResiduals = 6; ParameterDims = ceres::internal::ParameterDims<false, 4, 3, 4, 3>; Functor = RelativeRTError; T = double]’:

/usr/local/include/ceres/autodiff_cost_function.h:194:69: required from ‘bool ceres::AutoDiffCostFunction<CostFunctor, kNumResiduals, Ns>::Evaluate(const double* const*, double*, double**) const [with CostFunctor = RelativeRTError; int kNumResiduals = 6; int ...Ns = {4, 3, 4, 3}]’

/usr/local/include/ceres/autodiff_cost_function.h:184:8: required from here

/usr/local/include/ceres/internal/autodiff.h:313:9: error: no type named ‘Parameters’ in ‘class ceres::internal::ParameterDims<false, 4, 3, 4, 3>’

313 | using Parameters = typename ParameterDims::Parameters;

| ^~~~~~~~~~

/usr/local/include/ceres/internal/autodiff.h:346:47: error: no type named ‘Parameters’ in ‘class ceres::internal::ParameterDims<false, 4, 3, 4, 3>’

346 | Make1stOrderPerturbations<Parameters>::Apply(parameters,

| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~

347 | parameters_as_jets.data());

| ~~~~~~~~~~~~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/autodiff.h:355:39: error: no type named ‘Parameters’ in ‘class ceres::internal::ParameterDims<false, 4, 3, 4, 3>’

355 | Take1stOrderParts<Parameters>::Apply(

| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^

356 | num_outputs, residuals_as_jets.data(), jacobians);

| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/autodiff.h: In instantiation of ‘bool ceres::internal::AutoDifferentiate(const Functor&, const T* const*, int, T*, T**) [with int kNumResiduals = 3; ParameterDims = ceres::internal::ParameterDims<false, 3>; Functor = TError; T = double]’:

/usr/local/include/ceres/autodiff_cost_function.h:194:69: required from ‘bool ceres::AutoDiffCostFunction<CostFunctor, kNumResiduals, Ns>::Evaluate(const double* const*, double*, double**) const [with CostFunctor = TError; int kNumResiduals = 3; int ...Ns = {3}]’

/usr/local/include/ceres/autodiff_cost_function.h:184:8: required from here

/usr/local/include/ceres/internal/autodiff.h:313:9: error: no type named ‘Parameters’ in ‘class ceres::internal::ParameterDims<false, 3>’

313 | using Parameters = typename ParameterDims::Parameters;

| ^~~~~~~~~~

/usr/local/include/ceres/internal/autodiff.h:346:47: error: no type named ‘Parameters’ in ‘class ceres::internal::ParameterDims<false, 3>’

346 | Make1stOrderPerturbations<Parameters>::Apply(parameters,

| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~

347 | parameters_as_jets.data());

| ~~~~~~~~~~~~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/autodiff.h:355:39: error: no type named ‘Parameters’ in ‘class ceres::internal::ParameterDims<false, 3>’

355 | Take1stOrderParts<Parameters>::Apply(

| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^

356 | num_outputs, residuals_as_jets.data(), jacobians);

| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

In file included from /usr/local/include/ceres/internal/autodiff.h:152,

from /usr/local/include/ceres/autodiff_cost_function.h:130,

from /usr/local/include/ceres/ceres.h:37,

from /home/fastpx4/dai_ws/src/VINS-Fusion/global_fusion/src/globalOpt.h:20,

from /home/fastpx4/dai_ws/src/VINS-Fusion/global_fusion/src/globalOpt.cpp:12:

/usr/local/include/ceres/internal/variadic_evaluate.h: In instantiation of ‘bool ceres::internal::VariadicEvaluateImpl(const Functor&, const T* const*, T*, const void*) [with ParameterDims = ceres::internal::ParameterDims<false, 4, 3, 4, 3>; Functor = RelativeRTError; T = double]’:

/usr/local/include/ceres/internal/variadic_evaluate.h:100:45: required from ‘bool ceres::internal::VariadicEvaluate(const Functor&, const T* const*, T*) [with ParameterDims = ceres::internal::ParameterDims<false, 4, 3, 4, 3>; Functor = RelativeRTError; T = double]’

/usr/local/include/ceres/autodiff_cost_function.h:191:55: required from ‘bool ceres::AutoDiffCostFunction<CostFunctor, kNumResiduals, Ns>::Evaluate(const double* const*, double*, double**) const [with CostFunctor = RelativeRTError; int kNumResiduals = 6; int ...Ns = {4, 3, 4, 3}]’

/usr/local/include/ceres/autodiff_cost_function.h:184:8: required from here

/usr/local/include/ceres/internal/variadic_evaluate.h:74:52: error: ‘ParameterBlockIndices’ was not declared in this scope

74 | ParameterBlockIndices());

| ~~~~~~~~~~~~~~~~~~~~~^~

/usr/local/include/ceres/internal/variadic_evaluate.h: In instantiation of ‘bool ceres::internal::VariadicEvaluateImpl(const Functor&, const T* const*, T*, const void*) [with ParameterDims = ceres::internal::ParameterDims<false, 3>; Functor = TError; T = double]’:

/usr/local/include/ceres/internal/variadic_evaluate.h:100:45: required from ‘bool ceres::internal::VariadicEvaluate(const Functor&, const T* const*, T*) [with ParameterDims = ceres::internal::ParameterDims<false, 3>; Functor = TError; T = double]’

/usr/local/include/ceres/autodiff_cost_function.h:191:55: required from ‘bool ceres::AutoDiffCostFunction<CostFunctor, kNumResiduals, Ns>::Evaluate(const double* const*, double*, double**) const [with CostFunctor = TError; int kNumResiduals = 3; int ...Ns = {3}]’

/usr/local/include/ceres/autodiff_cost_function.h:184:8: required from here

/usr/local/include/ceres/internal/variadic_evaluate.h:74:52: error: ‘ParameterBlockIndices’ was not declared in this scope

make[2]: *** [VINS-Fusion/global_fusion/CMakeFiles/global_fusion_node.dir/build.make:76:VINS-Fusion/global_fusion/CMakeFiles/global_fusion_node.dir/src/globalOpt.cpp.o] 错误 1

make[2]: *** 正在等待未完成的任务....

In file included from /usr/local/include/ceres/internal/parameter_dims.h:37,

from /usr/local/include/ceres/internal/autodiff.h:151,

from /usr/local/include/ceres/autodiff_cost_function.h:130,

from /usr/local/include/ceres/ceres.h:37,

from /home/fastpx4/dai_ws/src/VINS-Fusion/global_fusion/src/globalOpt.h:20,

from /home/fastpx4/dai_ws/src/VINS-Fusion/global_fusion/src/globalOptNode.cpp:13:

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:64:21: error: ‘integer_sequence’ is not a member of ‘std’

64 | struct SumImpl<std::integer_sequence<T, N, Ns...>> {

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:64:21: error: ‘integer_sequence’ is not a member of ‘std’

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:64:46: error: wrong number of template arguments (3, should be 1)

64 | struct SumImpl<std::integer_sequence<T, N, Ns...>> {

| ^~~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:60:8: note: provided for ‘template<class Seq> struct ceres::internal::SumImpl’

60 | struct SumImpl;

| ^~~~~~~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:64:49: error: expected unqualified-id before ‘>’ token

64 | struct SumImpl<std::integer_sequence<T, N, Ns...>> {

| ^~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:70:21: error: ‘integer_sequence’ is not a member of ‘std’

70 | struct SumImpl<std::integer_sequence<T, N1, N2, Ns...>> {

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:70:21: error: ‘integer_sequence’ is not a member of ‘std’

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:70:51: error: wrong number of template arguments (4, should be 1)

70 | struct SumImpl<std::integer_sequence<T, N1, N2, Ns...>> {

| ^~~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:60:8: note: provided for ‘template<class Seq> struct ceres::internal::SumImpl’

60 | struct SumImpl;

| ^~~~~~~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:70:54: error: expected unqualified-id before ‘>’ token

70 | struct SumImpl<std::integer_sequence<T, N1, N2, Ns...>> {

| ^~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:77:21: error: ‘integer_sequence’ is not a member of ‘std’

77 | struct SumImpl<std::integer_sequence<T, N1, N2, N3, N4, Ns...>> {

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:77:21: error: ‘integer_sequence’ is not a member of ‘std’

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:77:59: error: wrong number of template arguments (6, should be 1)

77 | struct SumImpl<std::integer_sequence<T, N1, N2, N3, N4, Ns...>> {

| ^~~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:60:8: note: provided for ‘template<class Seq> struct ceres::internal::SumImpl’

60 | struct SumImpl;

| ^~~~~~~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:77:62: error: expected unqualified-id before ‘>’ token

77 | struct SumImpl<std::integer_sequence<T, N1, N2, N3, N4, Ns...>> {

| ^~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:84:21: error: ‘integer_sequence’ is not a member of ‘std’

84 | struct SumImpl<std::integer_sequence<T, N>> {

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:84:21: error: ‘integer_sequence’ is not a member of ‘std’

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:84:41: error: wrong number of template arguments (2, should be 1)

84 | struct SumImpl<std::integer_sequence<T, N>> {

| ^

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:60:8: note: provided for ‘template<class Seq> struct ceres::internal::SumImpl’

60 | struct SumImpl;

| ^~~~~~~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:84:42: error: expected unqualified-id before ‘>’ token

84 | struct SumImpl<std::integer_sequence<T, N>> {

| ^~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:90:21: error: ‘integer_sequence’ is not a member of ‘std’

90 | struct SumImpl<std::integer_sequence<T>> {

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:90:21: error: ‘integer_sequence’ is not a member of ‘std’

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:90:38: error: template argument 1 is invalid

90 | struct SumImpl<std::integer_sequence<T>> {

| ^

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:90:39: error: expected unqualified-id before ‘>’ token

90 | struct SumImpl<std::integer_sequence<T>> {

| ^~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:132:39: error: ‘integer_sequence’ is not a member of ‘std’

132 | struct ExclusiveScanImpl<T, Sum, std::integer_sequence<T, N, Ns...>,

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:132:39: error: ‘integer_sequence’ is not a member of ‘std’

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:132:67: error: template argument 3 is invalid

132 | struct ExclusiveScanImpl<T, Sum, std::integer_sequence<T, N, Ns...>,

| ^

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:132:67: error: type/value mismatch at argument 4 in template parameter list for ‘template<class T, T Sum, class SeqIn, class SeqOut> struct ceres::internal::ExclusiveScanImpl’

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:132:67: note: expected a type, got ‘N’

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:141:39: error: ‘integer_sequence’ is not a member of ‘std’

141 | struct ExclusiveScanImpl<T, Sum, std::integer_sequence<T>, SeqOut> {

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:141:39: error: ‘integer_sequence’ is not a member of ‘std’

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:141:57: error: wrong number of template arguments (3, should be 4)

141 | struct ExclusiveScanImpl<T, Sum, std::integer_sequence<T>, SeqOut> {

| ^

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:129:8: note: provided for ‘template<class T, T Sum, class SeqIn, class SeqOut> struct ceres::internal::ExclusiveScanImpl’

129 | struct ExclusiveScanImpl;

| ^~~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:155:53: error: ‘integer_sequence’ is not a member of ‘std’

155 | typename ExclusiveScanImpl<T, T(0), Seq, std::integer_sequence<T>>::Type;

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:155:53: error: ‘integer_sequence’ is not a member of ‘std’

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:155:70: error: template argument 4 is invalid

155 | typename ExclusiveScanImpl<T, T(0), Seq, std::integer_sequence<T>>::Type;

| ^

/usr/local/include/ceres/internal/integer_sequence_algorithm.h:155:16: error: expected nested-name-specifier

155 | typename ExclusiveScanImpl<T, T(0), Seq, std::integer_sequence<T>>::Type;

| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

In file included from /usr/local/include/ceres/internal/autodiff.h:151,

from /usr/local/include/ceres/autodiff_cost_function.h:130,

from /usr/local/include/ceres/ceres.h:37,

from /home/fastpx4/dai_ws/src/VINS-Fusion/global_fusion/src/globalOpt.h:20,

from /home/fastpx4/dai_ws/src/VINS-Fusion/global_fusion/src/globalOptNode.cpp:13:

/usr/local/include/ceres/internal/parameter_dims.h:44:55: error: ‘integer_sequence’ is not a member of ‘std’

44 | constexpr bool IsValidParameterDimensionSequence(std::integer_sequence<int>) {

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/parameter_dims.h:44:72: error: expected primary-expression before ‘int’

44 | constexpr bool IsValidParameterDimensionSequence(std::integer_sequence<int>) {

| ^~~

/usr/local/include/ceres/internal/parameter_dims.h:50:10: error: ‘template<int N, int ...Ts> constexpr const bool ceres::internal::IsValidParameterDimensionSequence’ redeclared as different kind of entity

50 | std::integer_sequence<int, N, Ts...>) {

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/parameter_dims.h:44:16: note: previous declaration ‘constexpr const bool ceres::internal::IsValidParameterDimensionSequence’

44 | constexpr bool IsValidParameterDimensionSequence(std::integer_sequence<int>) {

| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/parameter_dims.h:50:10: error: ‘integer_sequence’ is not a member of ‘std’

50 | std::integer_sequence<int, N, Ts...>) {

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/parameter_dims.h:50:27: error: expected primary-expression before ‘int’

50 | std::integer_sequence<int, N, Ts...>) {

| ^~~

/usr/local/include/ceres/internal/parameter_dims.h:50:40: error: expected ‘)’ before ‘>’ token

50 | std::integer_sequence<int, N, Ts...>) {

| ^

| )

/usr/local/include/ceres/internal/parameter_dims.h:49:49: note: to match this ‘(’

49 | constexpr bool IsValidParameterDimensionSequence(

| ^

/usr/local/include/ceres/internal/parameter_dims.h:69:27: error: ‘integer_sequence’ in namespace ‘std’ does not name a template type

69 | using Parameters = std::integer_sequence<int, Ns...>;

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/parameter_dims.h:74:41: error: there are no arguments to ‘Parameters’ that depend on a template parameter, so a declaration of ‘Parameters’ must be available [-fpermissive]

74 | IsValidParameterDimensionSequence(Parameters());

| ^~~~~~~~~~

/usr/local/include/ceres/internal/parameter_dims.h:74:41: note: (if you use ‘-fpermissive’, G++ will accept your code, but allowing the use of an undeclared name is deprecated)

/usr/local/include/ceres/internal/parameter_dims.h:85:16: error: ‘integer_sequence’ is not a member of ‘std’

85 | Sum<std::integer_sequence<int, Ns...>>::Value;

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/parameter_dims.h:85:16: error: ‘integer_sequence’ is not a member of ‘std’

/usr/local/include/ceres/internal/parameter_dims.h:85:43: error: template argument 1 is invalid

85 | Sum<std::integer_sequence<int, Ns...>>::Value;

| ^~

/usr/local/include/ceres/internal/parameter_dims.h:101:20: error: ‘std::integer_sequence’ has not been declared

101 | T* ptr, std::integer_sequence<int, Indices...>) {

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/parameter_dims.h:101:36: error: expected ‘,’ or ‘...’ before ‘<’ token

101 | T* ptr, std::integer_sequence<int, Indices...>) {

| ^

/usr/local/include/ceres/internal/parameter_dims.h: In static member function ‘static std::array<T*, ceres::internal::ParameterDims<IsDynamic, Ns>::kNumParameterBlocks> ceres::internal::ParameterDims<IsDynamic, Ns>::GetUnpackedParameters(T*)’:

/usr/local/include/ceres/internal/parameter_dims.h:94:35: error: ‘Parameters’ was not declared in this scope; did you mean ‘ParameterDims’?

94 | using Offsets = ExclusiveScan<Parameters>;

| ^~~~~~~~~~

| ParameterDims

/usr/local/include/ceres/internal/parameter_dims.h:94:45: error: template argument 1 is invalid

94 | using Offsets = ExclusiveScan<Parameters>;

| ^

/usr/local/include/ceres/internal/parameter_dims.h:95:39: error: there are no arguments to ‘Offsets’ that depend on a template parameter, so a declaration of ‘Offsets’ must be available [-fpermissive]

95 | return GetUnpackedParameters(ptr, Offsets());

| ^~~~~~~

In file included from /usr/local/include/ceres/internal/autodiff.h:152,

from /usr/local/include/ceres/autodiff_cost_function.h:130,

from /usr/local/include/ceres/ceres.h:37,

from /home/fastpx4/dai_ws/src/VINS-Fusion/global_fusion/src/globalOpt.h:20,

from /home/fastpx4/dai_ws/src/VINS-Fusion/global_fusion/src/globalOptNode.cpp:13:

/usr/local/include/ceres/internal/variadic_evaluate.h: At global scope:

/usr/local/include/ceres/internal/variadic_evaluate.h:51:39: error: ‘std::integer_sequence’ has not been declared

51 | std::integer_sequence<int, Indices...>) {

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/variadic_evaluate.h:51:55: error: expected ‘,’ or ‘...’ before ‘<’ token

51 | std::integer_sequence<int, Indices...>) {

| ^

/usr/local/include/ceres/internal/variadic_evaluate.h:62:39: error: ‘std::integer_sequence’ has not been declared

62 | std::integer_sequence<int>) {

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/variadic_evaluate.h:62:55: error: expected ‘,’ or ‘...’ before ‘<’ token

62 | std::integer_sequence<int>) {

| ^

/usr/local/include/ceres/internal/variadic_evaluate.h: In function ‘bool ceres::internal::VariadicEvaluateImpl(const Functor&, const T* const*, T*, const void*)’:

/usr/local/include/ceres/internal/variadic_evaluate.h:71:12: error: ‘make_integer_sequence’ in namespace ‘std’ does not name a template type

71 | std::make_integer_sequence<int, ParameterDims::kNumParameterBlocks>;

| ^~~~~~~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/variadic_evaluate.h:74:31: error: there are no arguments to ‘ParameterBlockIndices’ that depend on a template parameter, so a declaration of ‘ParameterBlockIndices’ must be available [-fpermissive]

74 | ParameterBlockIndices());

| ^~~~~~~~~~~~~~~~~~~~~

In file included from /usr/local/include/ceres/autodiff_cost_function.h:130,

from /usr/local/include/ceres/ceres.h:37,

from /home/fastpx4/dai_ws/src/VINS-Fusion/global_fusion/src/globalOpt.h:20,

from /home/fastpx4/dai_ws/src/VINS-Fusion/global_fusion/src/globalOptNode.cpp:13:

/usr/local/include/ceres/internal/autodiff.h: At global scope:

/usr/local/include/ceres/internal/autodiff.h:217:39: error: ‘integer_sequence’ is not a member of ‘std’

217 | struct Make1stOrderPerturbations<std::integer_sequence<int, N, Ns...>,

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/autodiff.h:217:39: error: ‘integer_sequence’ is not a member of ‘std’

/usr/local/include/ceres/internal/autodiff.h:217:69: error: template argument 1 is invalid

217 | struct Make1stOrderPerturbations<std::integer_sequence<int, N, Ns...>,

| ^

/usr/local/include/ceres/internal/autodiff.h:232:39: error: ‘integer_sequence’ is not a member of ‘std’

232 | struct Make1stOrderPerturbations<std::integer_sequence<int>, ParameterIdx, Total> {

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/autodiff.h:232:39: error: ‘integer_sequence’ is not a member of ‘std’

/usr/local/include/ceres/internal/autodiff.h:232:59: error: template argument 1 is invalid

232 | struct Make1stOrderPerturbations<std::integer_sequence<int>, ParameterIdx, Total> {

| ^

/usr/local/include/ceres/internal/autodiff.h:280:31: error: ‘integer_sequence’ is not a member of ‘std’

280 | struct Take1stOrderParts<std::integer_sequence<int, N, Ns...>,

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/autodiff.h:280:31: error: ‘integer_sequence’ is not a member of ‘std’

/usr/local/include/ceres/internal/autodiff.h:280:61: error: template argument 1 is invalid

280 | struct Take1stOrderParts<std::integer_sequence<int, N, Ns...>,

| ^

/usr/local/include/ceres/internal/autodiff.h:296:31: error: ‘integer_sequence’ is not a member of ‘std’

296 | struct Take1stOrderParts<std::integer_sequence<int>, ParameterIdx, Offset> {

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/autodiff.h:296:31: error: ‘integer_sequence’ is not a member of ‘std’

/usr/local/include/ceres/internal/autodiff.h:296:51: error: template argument 1 is invalid

296 | struct Take1stOrderParts<std::integer_sequence<int>, ParameterIdx, Offset> {

| ^

In file included from /usr/local/include/ceres/ceres.h:42,

from /home/fastpx4/dai_ws/src/VINS-Fusion/global_fusion/src/globalOpt.h:20,

from /home/fastpx4/dai_ws/src/VINS-Fusion/global_fusion/src/globalOptNode.cpp:13:

/usr/local/include/ceres/cost_function_to_functor.h:161:29: error: ‘std::integer_sequence’ has not been declared

161 | std::integer_sequence<int, Indices...>) {

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/cost_function_to_functor.h:161:45: error: expected ‘,’ or ‘...’ before ‘<’ token

161 | std::integer_sequence<int, Indices...>) {

| ^

/usr/local/include/ceres/cost_function_to_functor.h: In member function ‘bool ceres::CostFunctionToFunctor<kNumResiduals, Ns>::operator()(const T*, Ts* ...) const’:

/usr/local/include/ceres/cost_function_to_functor.h:147:14: error: ‘make_integer_sequence’ in namespace ‘std’ does not name a template type

147 | std::make_integer_sequence<int,

| ^~~~~~~~~~~~~~~~~~~~~

/usr/local/include/ceres/cost_function_to_functor.h:150:41: error: there are no arguments to ‘Indices’ that depend on a template parameter, so a declaration of ‘Indices’ must be available [-fpermissive]

150 | GetParameterPointers<T>(params, Indices());

| ^~~~~~~

In file included from /usr/local/include/ceres/dynamic_numeric_diff_cost_function.h:44,

from /usr/local/include/ceres/ceres.h:48,

from /home/fastpx4/dai_ws/src/VINS-Fusion/global_fusion/src/globalOpt.h:20,

from /home/fastpx4/dai_ws/src/VINS-Fusion/global_fusion/src/globalOptNode.cpp:13:

/usr/local/include/ceres/internal/numeric_diff.h: At global scope:

/usr/local/include/ceres/internal/numeric_diff.h:441:48: error: ‘integer_sequence’ is not a member of ‘std’

441 | std::integer_sequence<int, N, Ns...>,

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/numeric_diff.h:441:48: error: ‘integer_sequence’ is not a member of ‘std’

/usr/local/include/ceres/internal/numeric_diff.h:441:78: error: template argument 2 is invalid

441 | std::integer_sequence<int, N, Ns...>,

| ^

/usr/local/include/ceres/internal/numeric_diff.h:485:63: error: ‘integer_sequence’ is not a member of ‘std’

485 | struct EvaluateJacobianForParameterBlocks<ParameterDims, std::integer_sequence<int>,

| ^~~~~~~~~~~~~~~~

/usr/local/include/ceres/internal/numeric_diff.h:485:63: error: ‘integer_sequence’ is not a member of ‘std’

/usr/local/include/ceres/internal/numeric_diff.h:485:83: error: template argument 2 is invalid

485 | struct EvaluateJacobianForParameterBlocks<ParameterDims, std::integer_sequence<int>,

| ^

make[2]: *** [VINS-Fusion/global_fusion/CMakeFiles/global_fusion_node.dir/build.make:63:VINS-Fusion/global_fusion/CMakeFiles/global_fusion_node.dir/src/globalOptNode.cpp.o] 错误 1

make[1]: *** [CMakeFiles/Makefile2:1167:VINS-Fusion/global_fusion/CMakeFiles/global_fusion_node.dir/all] 错误 2

make: *** [Makefile:141:all] 错误 2

Invoking "make -j8 -l8" failed

2. 修改Cmakelist (四个文件夹)

1). 将c++11 修改为c++14

set(CMAKE_BUILD_TYPE "Release")

set(CMAKE_CXX_FLAGS "-std=c++14")

set(CMAKE_CXX_FLAGS_RELEASE "-O3 -fPIC")

2). 明确指定find_package中的OpenCV的版本:

#find_package(OpenCV REQUIRED)  

find_package(OpenCV 4 REQUIRED)

3. 添加头文件

1). camera_models>include>camodocal>chessboard>Chessboard.h中添加

#include <opencv2/imgproc/types_c.h>

#include <opencv2/calib3d/calib3d_c.h>

2). 在以下几个文件中添加:

     camera_models/include/camodocal/calib/CameraCalibration.h     loop_fusion/src/pose_graph.h     vins_estimator/src/featureTracker/feature_tracker.h     loop_fusion/src/ThirdParty/DVision/BRIEF.h

#include <opencv2/imgproc/types_c.h>

#include <opencv2/imgproc/imgproc_c.h>

4. 仍然有一部分错误:

error: ‘CV_LOAD_IMAGE_GRAYSCALE’ was not declared in this scope

122 | imLeft = cv::imread(leftImagePath, CV_LOAD_IMAGE_GRAYSCALE );

| ^~~~~~~~~~~~~~~~~~~~~~~

/home/fastpx4/dai_ws/src/VINS-Fusion/vins_estimator/src/KITTIOdomTest.cpp: In function ‘int main(int, char**)’:

/home/fastpx4/dai_ws/src/VINS-Fusion/vins_estimator/src/KITTIOdomTest.cpp:95:39: error: ‘CV_LOAD_IMAGE_GRAYSCALE’ was not declared in this scope

95 | imLeft = cv::imread(leftImagePath, CV_LOAD_IMAGE_GRAYSCALE );

|

处理:将报错文件中的 CV_LOAD_IMAGE_GRAYSCALE修改为cv::IMREAD_GRAYSCALE

 1). 修改报错文件:VINS-Fusion/vins_estimator/src/KITTIGPSTest.cpp

122 imLeft = cv::imread(leftImagePath, CV_LOAD_IMAGE_GRAYSCALE );

123 imRight = cv::imread(rightImagePath, CV_LOAD_IMAGE_GRAYSCALE );

修改后:

imLeft = cv::imread(leftImagePath, cv::IMREAD_GRAYSCALE );

imRight = cv::imread(rightImagePath, cv::IMREAD_GRAYSCALE );

 2). 修改文件:VINS-Fusion/vins_estimator/src/KITTIOdomTest.cpp

imLeft = cv::imread(leftImagePath, CV_LOAD_IMAGE_GRAYSCALE );

sensor_msgs::ImagePtr imLeftMsg = cv_bridge::CvImage(std_msgs::Header(), "mono8", imLeft).toImageMsg();

imLeftMsg->header.stamp = ros::Time(imageTimeList[i]);

pubLeftImage.publish(imLeftMsg);

imRight = cv::imread(rightImagePath, CV_LOAD_IMAGE_GRAYSCALE );

sensor_msgs::ImagePtr imRightMsg = cv_bridge::CvImage(std_msgs::Header(), "mono8", imRight).toImageMsg();

imRightMsg->header.stamp = ros::Time(imageTimeList[i]);

pubRightImage.publish(imRightMsg);

修改后:

imLeft = cv::imread(leftImagePath, cv::IMREAD_GRAYSCALE );

sensor_msgs::ImagePtr imLeftMsg = cv_bridge::CvImage(std_msgs::Header(), "mono8", imLeft).toImageMsg();

imLeftMsg->header.stamp = ros::Time(imageTimeList[i]);

pubLeftImage.publish(imLeftMsg);

imRight = cv::imread(rightImagePath, cv::IMREAD_GRAYSCALE );

sensor_msgs::ImagePtr imRightMsg = cv_bridge::CvImage(std_msgs::Header(), "mono8", imRight).toImageMsg();

imRightMsg->header.stamp = ros::Time(imageTimeList[i]);

pubRightImage.publish(imRightMsg);

 二、编译通过!测试运行:

source devel/setup.bash

1. 运行数据集Euroc:

roslaunch vins vins_rviz.launch

rosrun vins vins_node src/VINS-Fusion/config/euroc/euroc_stereo_imu_config.yaml

rosrun loop_fusion loop_fusion_node src/VINS-Fusion/config/euroc/euroc_stereo_imu_config.yaml

cd ../data/

rosbag play MH_01_easy.bag

运行成功

2. 运行OAK-D 双目相机:(教程网址)

1.4. ubuntu 18.04 OAK-D系列相机 运行VINS-Fusion 双目+IMU — DepthAI Docs 0.3.0.0 documentation

icon-default.png?t=N7T8

https://docs.oakchina.cn/en/latest/pages/HighLevel/ros/ros_VINS_Fusion.html按照官网教程安装相机驱动,可以使用系统自带的opencv4.2版本进行编译。

<code>roslaunch depthai_examples stereo_inertial_node.launch enableRviz:=false depth_aligned:=false

roslaunch vins vins_rviz.launch

rosrun vins vins_node ~/dai_ws/src/VINS-Fusion/config/oak_d/config.yaml

(optional) rosrun loop_fusion loop_fusion_node ~/dai_ws/src/VINS-Fusion/config/oak_d/config.yaml

 注意:直接使用官网配置文件可能导致无法运行

在官网教程中使用的是OAK-D-S2相机,我使用的是OAK-D-PRO相机,配置参数有一些区别,需要对配置文件进行针对性的修改(在left.yaml和right.yaml文件中修改相机畸变参数),才可以正常运行VINS-Fusion程序。

1. 启动相机驱动,确认相机正常运行:

roslaunch depthai_examples stereo_inertial_node.launch depth_aligned:=false

*depth_aligned:=false 此参数控制是否输出深度映射图像,这里是不输出

2.查看话题输出

rostopic list

/clicked_point

/initialpose

/joint_states

/move_base_simple/goal

/nodelet_manager/bond

/rosout

/rosout_agg

/stereo_inertial_publisher/right/camera_info

/stereo_inertial_publisher/right/image_rect

/stereo_inertial_publisher/stereo/camera_info

/stereo_inertial_publisher/stereo/depth

/stereo_inertial_publisher/stereo/image

/stereo_inertial_publisher/stereo/image/compressed

/stereo_inertial_publisher/stereo/image/compressed/parameter_descriptions

/stereo_inertial_publisher/stereo/image/compressed/parameter_updates

/stereo_inertial_publisher/stereo/image/compressedDepth

/stereo_inertial_publisher/stereo/image/compressedDepth/parameter_descriptions

/stereo_inertial_publisher/stereo/image/compressedDepth/parameter_updates

/stereo_inertial_publisher/stereo/image/theora

/stereo_inertial_publisher/stereo/image/theora/parameter_descriptions

/stereo_inertial_publisher/stereo/image/theora/parameter_updates

/stereo_inertial_publisher/stereo/points

/tf

/tf_static

3. 输出话题数据,并将参数写入left.yaml和right.yaml中:

rostopic echo /stereo_inertial_publisher/left/camera_info

header:

seq: 7

stamp:

secs: 1711676690

nsecs: 777232861

frame_id: "oak_left_camera_optical_frame"

height: 720

width: 1280

distortion_model: "rational_polynomial"

D: [-11.679576873779297, 105.025146484375, -0.0006934021948836744, 0.0018576113507151604, -51.34315490722656, -11.695950508117676, 104.31090545654297, -49.294837951660156]

K: [798.2916259765625, 0.0, 651.572265625, 0.0, 798.2916259765625, 361.30743408203125, 0.0, 0.0, 1.0]

R: [0.9997826814651489, 0.002714010654017329, -0.020668860524892807, -0.00276787206530571, 0.999992847442627, -0.00257775466889143, 0.020661717280745506, 0.002634403295814991, 0.9997830390930176]

P: [799.2300415039062, 0.0, 654.2191772460938, 59.83273947490379, 0.0, 799.2300415039062, 359.2149658203125, 0.0, 0.0, 0.0, 1.0, 0.0]

binning_x: 0

binning_y: 0

roi:

x_offset: 0

y_offset: 0

height: 0

width: 0

do_rectify: False

---

将K矩阵对应填入文件left.yaml中

%YAML:1.0

---

model_type: PINHOLE

camera_name: camera

image_width: 640

image_height: 400

distortion_parameters:

k1: 0.011099142353676499

k2: -0.05769482092275897

p1: -0.0009757653113701839

p2: 0.0025548857914714745

projection_parameters:

fx: 798.2916259765625

fy: 798.2916259765625

cx: 651.572265625

cy: 361.30743408203125

rostopic echo /stereo_inertial_publisher/right/camera_info

header:

seq: 291

stamp:

secs: 1711676701

nsecs: 243829779

frame_id: "oak_right_camera_optical_frame"

height: 720

width: 1280

distortion_model: "rational_polynomial"

D: [-6.891867160797119, 81.7864990234375, -0.00016954877355601639, 0.0002883463748730719, -11.735593795776367, -6.942117691040039, 81.16423034667969, -9.775546073913574]

K: [799.2300415039062, 0.0, 654.2191772460938, 0.0, 799.2300415039062, 359.2149658203125, 0.0, 0.0, 1.0]

R: [0.999980092048645, -0.0040285540744662285, 0.004851967561990023, 0.004015895072370768, 0.9999884963035583, 0.002615976147353649, -0.0048624505288898945, -0.0025964390952140093, 0.9999848008155823]

P: [799.2300415039062, 0.0, 654.2191772460938, 0.0, 0.0, 799.2300415039062, 359.2149658203125, 0.0, 0.0, 0.0, 1.0, 0.0]

binning_x: 0

binning_y: 0

roi:

x_offset: 0

y_offset: 0

height: 0

width: 0

do_rectify: False

---

%YAML:1.0

---

model_type: PINHOLE

camera_name: camera

image_width: 640

image_height: 400

distortion_parameters:

k1: 0.005536021298200547

k2: -0.048229113205249675

p1: -0.0002985290327403832

p2: 0.0037187807087799125

projection_parameters:

fx: 799.2300415039062

fy: 799.2300415039062

cx: 654.2191772460938

cy: 359.2149658203125

 修改完成后,重新启动相机驱动及VIns-Fusion程序,即可成功运行。

*补充:

在文件(depthai-ros/depthai_examples/launch/stereo_inertial_node.launch)中可以发现,参数depth_aligned会决定输出话题是左右目图像还是深度图映射。

<code> <node if="$(eval arg('depth_aligned') == true)" pkg="nodelet" type="nodelet" name="depth_image_to_rgb_pointcloud"code>

args="load depth_image_proc/point_cloud_xyzrgb nodelet_manager">code>

<param name="queue_size" value="10"/>code>

<remap from="rgb/camera_info" to="/stereo_inertial_publisher/color/camera_info"/>code>

<remap from="rgb/image_rect_color" to="/stereo_inertial_publisher/color/image"/>code>

<remap from="depth_registered/image_rect" to="/stereo_inertial_publisher/stereo/image"/> code>

<remap from="depth_registered/points" to="/stereo_inertial_publisher/stereo/points"/>code>

</node>

<node if="$(eval arg('depth_aligned') == false)" pkg="nodelet" type="nodelet" name="depth_image_to_pointcloud"code>

args="load depth_image_proc/point_cloud_xyz nodelet_manager">code>

<param name="queue_size" value="10"/>code>

<remap from="camera_info" to="/stereo_inertial_publisher/stereo/camera_info"/>code>

<remap from="image_rect" to="/stereo_inertial_publisher/stereo/image"/> code>

<remap from="points" to="/stereo_inertial_publisher/stereo/points"/>code>

</node>

如果启动相机驱动时不关闭这个参数,话题输出会不同。

/clicked_point

/initialpose

/joint_states

/move_base_simple/goal

/nodelet_manager/bond

/rosout

/rosout_agg

/stereo_inertial_publisher/color/camera_info

/stereo_inertial_publisher/color/image

/stereo_inertial_publisher/color/image/compressed

/stereo_inertial_publisher/color/image/compressed/parameter_descriptions

/stereo_inertial_publisher/color/image/compressed/parameter_updates

/stereo_inertial_publisher/color/image/compressedDepth

/stereo_inertial_publisher/color/image/compressedDepth/parameter_descriptions

/stereo_inertial_publisher/color/image/compressedDepth/parameter_updates

/stereo_inertial_publisher/color/image/theora

/stereo_inertial_publisher/color/image/theora/parameter_descriptions

/stereo_inertial_publisher/color/image/theora/parameter_updates

/stereo_inertial_publisher/color/preview/camera_info

/stereo_inertial_publisher/color/preview/image

/stereo_inertial_publisher/color/preview/image/compressed

/stereo_inertial_publisher/color/preview/image/compressed/parameter_descriptions

/stereo_inertial_publisher/color/preview/image/compressed/parameter_updates

/stereo_inertial_publisher/color/preview/image/compressedDepth

/stereo_inertial_publisher/color/preview/image/compressedDepth/parameter_descriptions

/stereo_inertial_publisher/color/preview/image/compressedDepth/parameter_updates

/stereo_inertial_publisher/color/preview/image/theora

/stereo_inertial_publisher/color/preview/image/theora/parameter_descriptions

/stereo_inertial_publisher/color/preview/image/theora/parameter_updates

/stereo_inertial_publisher/color/yolov4_Spatial_detections

/stereo_inertial_publisher/imu

/stereo_inertial_publisher/stereo/camera_info

/stereo_inertial_publisher/stereo/depth

/stereo_inertial_publisher/stereo/depth/compressed

/stereo_inertial_publisher/stereo/depth/compressed/parameter_descriptions

/stereo_inertial_publisher/stereo/depth/compressed/parameter_updates

/stereo_inertial_publisher/stereo/depth/compressedDepth

/stereo_inertial_publisher/stereo/depth/compressedDepth/parameter_descriptions

/stereo_inertial_publisher/stereo/depth/compressedDepth/parameter_updates

/stereo_inertial_publisher/stereo/depth/theora

/stereo_inertial_publisher/stereo/depth/theora/parameter_descriptions

/stereo_inertial_publisher/stereo/depth/theora/parameter_updates

/stereo_inertial_publisher/stereo/image

/stereo_inertial_publisher/stereo/image/compressed

/stereo_inertial_publisher/stereo/image/compressed/parameter_descriptions

/stereo_inertial_publisher/stereo/image/compressed/parameter_updates

/stereo_inertial_publisher/stereo/image/compressedDepth

/stereo_inertial_publisher/stereo/image/compressedDepth/parameter_descriptions

/stereo_inertial_publisher/stereo/image/compressedDepth/parameter_updates

/stereo_inertial_publisher/stereo/image/theora

/stereo_inertial_publisher/stereo/image/theora/parameter_descriptions

/stereo_inertial_publisher/stereo/image/theora/parameter_updates

/stereo_inertial_publisher/stereo/points

/tf

/tf_static

3. 运行realsense_D435i 双目相机:

 由于OAK-D相机暂时无法成功运行ego-planner程序,先找了一个原框架适配的realsense的相机。

1. 启动相机驱动

roslaunch realsense2_camera rs_d435_camera_with_model.launch

2. 打开Rviz 或 rqt_image_view查看,确认有话题输出

        /camera/depth/image_rect_raw(深度图)、

        /camera/infra1/image_rect_raw(左目图像)、

        /camera/infra2/image_rect_raw(右目图像)

3. 修改相机标定数据,如同上述步骤,此处省略。

三、VIns-Fusion配置参数矫正

参考浙江大学Fast-Lab教程

【完结】从0制作自主空中机器人 | 开源 | 浙江大学Fast-Lab_哔哩哔哩_bilibili

icon-default.png?t=N7T8

https://www.bilibili.com/video/BV1WZ4y167me/?spm_id_from=333.1007.top_right_bar_window_custom_collection.content.click&vd_source=7efc9d03142aa088952c10884de6dadfZJU-FAST-Lab/Fast-Drone-250: hardware and software design of the 250mm autonomous drone (github.com)

icon-default.png?t=N7T8

https://github.com/ZJU-FAST-Lab/Fast-Drone-250

1)启动Vins程序,查看输出话题;

<code>/clicked_point

/feature_tracker/feature

/globalEstimator/car_model

/globalEstimator/global_path

/initialpose

/joint_states

/loop_fusion/base_path

/loop_fusion/camera_pose_visual

/loop_fusion/margin_cloud_loop_rect

/loop_fusion/pose_graph

/move_base_simple/goal

/nodelet_manager/bond

/rosout

/rosout_agg

/stereo_inertial_publisher/imu

/stereo_inertial_publisher/left/camera_info

/stereo_inertial_publisher/left/image_rect

/stereo_inertial_publisher/left/image_rect/compressed

/stereo_inertial_publisher/left/image_rect/compressed/parameter_descriptions

/stereo_inertial_publisher/left/image_rect/compressed/parameter_updates

/stereo_inertial_publisher/left/image_rect/compressedDepth

/stereo_inertial_publisher/left/image_rect/compressedDepth/parameter_descriptions

/stereo_inertial_publisher/left/image_rect/compressedDepth/parameter_updates

/stereo_inertial_publisher/left/image_rect/theora

/stereo_inertial_publisher/left/image_rect/theora/parameter_descriptions

/stereo_inertial_publisher/left/image_rect/theora/parameter_updates

/stereo_inertial_publisher/right/camera_info

/stereo_inertial_publisher/right/image_rect

/stereo_inertial_publisher/right/image_rect/compressed

/stereo_inertial_publisher/right/image_rect/compressed/parameter_descriptions

/stereo_inertial_publisher/right/image_rect/compressed/parameter_updates

/stereo_inertial_publisher/right/image_rect/compressedDepth

/stereo_inertial_publisher/right/image_rect/compressedDepth/parameter_descriptions

/stereo_inertial_publisher/right/image_rect/compressedDepth/parameter_updates

/stereo_inertial_publisher/right/image_rect/theora

/stereo_inertial_publisher/right/image_rect/theora/parameter_descriptions

/stereo_inertial_publisher/right/image_rect/theora/parameter_updates

/stereo_inertial_publisher/stereo/camera_info

/stereo_inertial_publisher/stereo/depth

/stereo_inertial_publisher/stereo/depth/compressed

/stereo_inertial_publisher/stereo/depth/compressed/parameter_descriptions

/stereo_inertial_publisher/stereo/depth/compressed/parameter_updates

/stereo_inertial_publisher/stereo/depth/compressedDepth

/stereo_inertial_publisher/stereo/depth/compressedDepth/parameter_descriptions

/stereo_inertial_publisher/stereo/depth/compressedDepth/parameter_updates

/stereo_inertial_publisher/stereo/depth/theora

/stereo_inertial_publisher/stereo/depth/theora/parameter_descriptions

/stereo_inertial_publisher/stereo/depth/theora/parameter_updates

/stereo_inertial_publisher/stereo/image

/stereo_inertial_publisher/stereo/image/compressed

/stereo_inertial_publisher/stereo/image/compressed/parameter_descriptions

/stereo_inertial_publisher/stereo/image/compressed/parameter_updates

/stereo_inertial_publisher/stereo/image/compressedDepth

/stereo_inertial_publisher/stereo/image/compressedDepth/parameter_descriptions

/stereo_inertial_publisher/stereo/image/compressedDepth/parameter_updates

/stereo_inertial_publisher/stereo/image/theora

/stereo_inertial_publisher/stereo/image/theora/parameter_descriptions

/stereo_inertial_publisher/stereo/image/theora/parameter_updates

/stereo_inertial_publisher/stereo/points

/tf

/tf_static

/vins_cam_switch

/vins_fusion/camera_pose

/vins_fusion/camera_pose_visual

/vins_fusion/extrinsic

/vins_fusion/image_track

/vins_fusion/imu_propagate

/vins_fusion/key_poses

/vins_fusion/keyframe_point

/vins_fusion/keyframe_pose

/vins_fusion/margin_cloud

/vins_fusion/odometry

/vins_fusion/path

/vins_fusion/point_cloud

/vins_imu_switch

/vins_restart

 2)终端输入,查看程序输出position的漂移情况

rostopic echo /vins_fusion/imu_propagate

header:

seq: 30359

stamp:

secs: 1711678556

nsecs: 905394554

frame_id: "world"

child_frame_id: ''

pose:

pose:

position:

x: -0.0018558396542408368

y: 0.0006243309409628052

z: 0.002800527454180711

orientation:

x: 0.8243210546283225

y: 0.009086388930822076

z: -0.5658114242209056

w: 0.01642161438402998

covariance: [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]

twist:

twist:

linear:

x: 0.0003866194433702027

y: 0.0008024273412679251

z: 0.00022701032056310962

angular:

x: 0.0

y: 0.0

z: 0.0

covariance: [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]

---

3)(将相机与板载计算机固定在无人机上之后)抱起无人机在场地内缓慢走动,在vins的输出文件夹下打开extrinsic_parameter.txt

      将文件内容替换在config.yaml文件中对应位置。

%YAML:1.0

#common parameters

#support: 1 imu 1 cam; 1 imu 2 cam: 2 cam;

imu: 1

num_of_cam: 2

#imu_topic: "/mavros/imu/data_raw"

imu_topic: "/stereo_inertial_publisher/imu"

#/stereo_inertial_publisher/imu

image0_topic: "/stereo_inertial_publisher/left/image_rect"

image1_topic: "/stereo_inertial_publisher/right/image_rect"

output_path: "/home/fastpx4/abc/vins_output"

cam0_calib: "left.yaml"

cam1_calib: "right.yaml"

image_width: 640

#640

image_height: 400

#400

# Extrinsic parameter between IMU and Camera.

estimate_extrinsic: 1 # 0 Have an accurate extrinsic parameters. We will trust the following imu^R_cam, imu^T_cam, don't change it.

# 1 Have an initial guess about extrinsic parameters. We will optimize around your initial guess.

body_T_cam0: !!opencv-matrix

rows: 4

cols: 4

dt: d

data: [ 1.1582253353211946e-02, 9.9992842406279825e-01,

2.9996930761942851e-03, -1.5936022004130995e-02,

9.9992731716619121e-01, -1.1572090774962351e-02,

-3.3833563099195255e-03, -8.4883987362590493e-02,

-3.3484014224460091e-03, 3.0386619399666269e-03,

-9.9998977731851335e-01, -3.4809993558054787e-02, 0., 0., 0., 1. ]

body_T_cam1: !!opencv-matrix

rows: 4

cols: 4

dt: d

data: [ 1.1352436793192977e-02, 9.9993544249239663e-01,

4.8272807242477774e-04, -1.5103709396341888e-02,

9.9993074772788504e-01, -1.1350883268602896e-02,

-3.1076031560835337e-03, -9.5661037692648648e-03,

-3.1019231469685844e-03, 5.1797351081671458e-04,

-9.9999505487588936e-01, -3.3604401456376665e-02, 0., 0., 0., 1. ]

#Multiple thread support

multiple_thread: 1

#feature traker paprameters

max_cnt: 130 # max feature number in feature tracking

min_dist: 30 # min distance between two features

freq: 0 # frequence (Hz) of publish tracking result. At least 10Hz for good estimation. If set 0, the frequence will be same as raw image

F_threshold: 1.0 # ransac threshold (pixel)

show_track: 1 # publish tracking image as topic

flow_back: 1 # perform forward and backward optical flow to improve feature tracking accuracy

#optimization parameters

max_solver_time: 0.04 # max solver itration time (ms), to guarantee real time

max_num_iterations: 8 # max solver itrations, to guarantee real time

keyframe_parallax: 10.0 # keyframe selection threshold (pixel)

#imu parameters The more accurate parameters you provide, the better performance

acc_n: 0.1 # accelerometer measurement noise standard deviation.

gyr_n: 0.01 # gyroscope measurement noise standard deviation.

acc_w: 0.001 # accelerometer bias random work noise standard deviation.

gyr_w: 0.0001 # gyroscope bias random work noise standard deviation.

g_norm: 9.81007 # gravity magnitude

#unsynchronization parameters

estimate_td: 1 # online estimate time offset between camera and imu

td: 0.0 # initial value of time offset. unit: s. readed image clock + td = real image clock (IMU clock)

#loop closure parameters

load_previous_pose_graph: 0 # load and reuse previous pose graph; load from 'pose_graph_save_path'

pose_graph_save_path: "/home/fastpx4/abc/vins_output" # save and load path

save_image: 1 # save image in pose graph for visualization prupose; you can close this function by setting 0

*重复上述过程,覆盖配置参数,直到话题中位置输出收敛到满意程度。

 四、ego-planner导航

在浙大的Fast-Drone-250项目中,ego_planner相关的代码以及启动launch文件位置在src/planner/plan_manage。

1. 启动文件(single_run_in_exp.launch)

        1)修改深度图像话题,对应上述D435相机驱动位置的话题

<arg name="depth_topic" value="/camera/depth/image_rect_raw"/>code>

        2)修改相机参数

<arg name="cx" value="319.7215270996094"/>code>

<arg name="cy" value="239.18408203125"/>code>

<arg name="fx" value="385.1890869140625"/>code>

<arg name="fy" value="385.1890869140625"/>code>

        3)设置地图尺寸及vins里程计话题

<arg name="map_size_x" value="30"/>code>

<arg name="map_size_y" value="30"/>code>

<arg name="map_size_z" value="3.0"/>code>

<arg name="odom_topic" value="/vins_fusion/imu_propagate"/>code>

2. 参数配置文件(advanced_param_exp.xml)

        根据自己的无人机设定配置参数

<param name="grid_map/resolution" value="0.2" /> <!-- 分辨率 -->code>

<param name="grid_map/map_size_x" value="$(arg map_size_x_)" /> code>

<param name="grid_map/map_size_y" value="$(arg map_size_y_)" /> code>

<param name="grid_map/map_size_z" value="$(arg map_size_z_)" /> code>

<param name="grid_map/local_update_range_x" value="5.5" /> code>

<param name="grid_map/local_update_range_y" value="5.5" /> code>

<param name="grid_map/local_update_range_z" value="4.5" /> code>

<param name="grid_map/obstacles_inflation" value="0.4" /> <!-- 膨胀系数,最大以分辨率膨胀4倍-->code>

<param name="grid_map/local_map_margin" value="10"/>code>

<!-- 地面高度 -->

<param name="grid_map/ground_height" value="-0.01"/>code>

测试ego-planner建图效果

启动两个终端,分别输入:

roslaunch ego_planner single_run_in_exp.launch

roslaunch ego_planner rviz.launch

即可看到:(如果没有障碍物,检查深度图话题是否正常输出)

为方便启动,编写脚本,一次启动!

<code>source devel/setup.bash

roslaunch realsense2_camera rs_d435_camera_with_model.launch & sleep 8;

roslaunch mavros px4.launch & sleep 5;

roslaunch vins fast_drone_250.launch & sleep 3;

roslaunch ego_planner single_run_in_exp.launch & sleep 2 ;

roslaunch ego_planner rviz.launch ;

wait;

3. 自动起飞控制(src/realflight_modules/px4ctrl/launch/run_ctrl.launch)

在px4ctrl/config/ctrl_param_fpv.yaml文件下修改无人机重量、及自动起飞配置,悬停油门水平,其他参数可自行摸索,不需要重点关注。

#########################################################################

# ⭐⭐⭐ Set parameters carefully for satisfied performance! ⭐⭐⭐ #

#########################################################################

mass : 3.8 # kg

gra : 9.81

pose_solver : 1 # 0:From ZhepeiWang (drag & less singular) 1:From ZhepeiWang, 2:From rotor-drag

ctrl_freq_max : 100.0

use_bodyrate_ctrl: false

max_manual_vel: 1.0

max_angle: 30 # Attitude angle limit in degree. A negative value means no limit.

low_voltage: 13.2 # 4S battery

rc_reverse: # *通道反向

roll: false

pitch: false

yaw: false

throttle: false

auto_takeoff_land:

enable: true

enable_auto_arm: true

no_RC: false

takeoff_height: 1.0 # m起飞高度

takeoff_land_speed: 0.3 # m/s速度

thrust_model: # The model that maps thrust signal u(0~1) to real thrust force F(Unit:N): F=K1*Voltage^K2*(K3*u^2+(1-K3)*u).

print_value: true # display the value of “thr_scale_compensate” or “hover_percentage” during thrust model estimating.

accurate_thrust_model: false # This can always enabled if don't require accurate control performance :-)

# accurate thrust mapping parameters

K1: 0.7583 # Needs precise calibration!

K2: 1.6942 # Needs precise calibration!

K3: 0.6786 # Needs precise calibration! K3 equals THR_MDL_FAC in https://docs.px4.io/master/en/config_mc/pid_tuning_guide_multicopter.html.

# approximate thrust mapping parameters

hover_percentage: 0.50 # 悬停油门水平 Thrust percentage in Stabilize/Arco mode # *

gain:

# Cascade PID controller. Recommend to read the code.

Kp0: 1.5

Kp1: 1.5

Kp2: 1.5

Kv0: 1.5

Kv1: 1.5

Kv2: 1.5

# ↓↓↓ No use now --

Kvi0: 0.0

Kvi1: 0.0

Kvi2: 0.0

Kvd0: 0.0

Kvd1: 0.0

Kvd2: 0.0

# ↓↓↓ Only used in rate control mode.

KAngR: 20.0

KAngP: 20.0

KAngY: 20.0

rotor_drag:

x: 0.0 # The reduced acceleration on each axis caused by rotor drag. Unit:(m*s^-2)/(m*s^-1).

y: 0.0 # Same as above

z: 0.0 # Same as above

k_thrust_horz: 0.0 # Set to 0 recommended... --

#超时阈值

msg_timeout:

odom: 0.5

rc: 0.5

cmd: 0.5

imu: 0.5

bat: 0.5

4. 自主导航飞行(启动顺序)

1)启动双目相机

2)连接飞控

3)启动vins_fusion节点

4)启动ego-planner节点

5)启动ego可视化界面

source devel/setup.bash

roslaunch realsense2_camera rs_d435_camera_with_model.launch & sleep 8;

roslaunch mavros px4.launch & sleep 5;

roslaunch vins fast_drone_250.launch & sleep 3;

roslaunch ego_planner single_run_in_exp.launch & sleep 2 ;

roslaunch ego_planner rviz.launch ;

wait;

6)在原地抱着飞机转几圈,确认vins里程计不飘

7)启动无人机控制

roslaunch px4ctrl run_ctrl.launch

8)自动起飞(运行起飞脚本)

首先,确定打开飞机遥控器,进入板载模式,解锁。(可能不同的遥控器要求不同)

sh shfiles/takeoff.sh

起飞拒绝,可能原因,vins里程计漂移较大,遥控器模式不正确,具体可查看源码文件

(src/planner/plan_manage/src/ego_replan_fsm.cpp)

9)设定导航目标

选中rviz界面,按下G键,在地图中点击,设定导航目标。



声明

本文内容仅代表作者观点,或转载于其他网站,本站不以此文作为商业用途
如有涉及侵权,请联系本站进行删除
转载本站原创文章,请注明来源及作者。