Error in compiling dvo_core

Question

Error in compiling dvo_core

ple13 opened this issue 8 years ago · 12 comments

I'm trying to compile dvo_slam on my machine: Ubuntu 12.04.5 / ros fuerte. The compilation terminates when dvo_core is compiled.

In summary, I notice that there is an error relating to conversion between different data types.

In dense_tracking.cpp, line 263 causes the compilation error:
Eigen::Affine3f transformf;
transformf = estimate().matrix().cast();

I'm thinking about passing data from estimate().matrix() to transformf manually. However I guess there should be a better way to handle the issue.

Below is the error message:

{-------------------------------------------------------------------------------
/home/panda/Desktop/slam_baselines/dvo_slam/sophus/include/sophus/se3.hpp: In instantiation of ‘Eigen::Matrix<typename Eigen::internal::traits::Scalar, 3, 3> Sophus::SE3GroupBase::rotationMatrix() const [with Derived = Sophus::SE3Group; typename Eigen::internal::traits::Scalar = double]’:
/home/panda/Desktop/slam_baselines/dvo_slam/sophus/include/sophus/se3.hpp:227:59: required from ‘Sophus::SE3GroupBase::Transformation Sophus::SE3GroupBase::matrix() const [with Derived = Sophus::SE3Group; Sophus::SE3GroupBase::Transformation = Eigen::Matrix<double, 4, 4>; typename Eigen::internal::traits::Scalar = double]’
/home/panda/Desktop/slam_baselines/dvo_slam/dvo_core/src/dense_tracking.cpp:263:40: required from here
/home/panda/Desktop/slam_baselines/dvo_slam/sophus/include/sophus/se3.hpp:311:25: error: could not convert ‘((const Sophus::SO3GroupBaseSophus::SO3Group)(& Sophus::SE3GroupBase::so3Sophus::SE3Group()))->Sophus::SO3GroupBase::matrixSophus::SO3Group()’ from ‘Sophus::SO3GroupBaseSophus::SO3Group::Transformation {aka int}’ to ‘Eigen::Matrix<double, 3, 3>’
return so3().matrix();
^
In file included from /home/panda/Desktop/slam_baselines/dvo_slam/sophus/include/sophus/se3.hpp:27:0,
from /home/panda/Desktop/slam_baselines/dvo_slam/dvo_core/src/dense_tracking.cpp:27:
/home/panda/Desktop/slam_baselines/dvo_slam/sophus/include/sophus/so3.hpp: In instantiation of ‘static Sophus::SO3GroupBase::Tangent Sophus::SO3GroupBase::logAndTheta(const int&, int) [with Derived = Sophus::SO3Group; Sophus::SO3GroupBase::Tangent = int]’:
/home/panda/Desktop/slam_baselines/dvo_slam/sophus/include/sophus/se3.hpp:611:56: required from ‘static Sophus::SE3GroupBase::Tangent Sophus::SE3GroupBase::log(const Sophus::SE3Group<typename Eigen::internal::traits::Scalar>&) [with Derived = Sophus::SE3Group; Sophus::SE3GroupBase::Tangent = Eigen::Matrix<double, 6, 1>; typename Eigen::internal::traits::Scalar = double]’
/home/panda/Desktop/slam_baselines/dvo_slam/sophus/include/sophus/se3.hpp:211:21: required from ‘Sophus::SE3GroupBase::Tangent Sophus::SE3GroupBase::log() const [with Derived = Sophus::SE3Group; Sophus::SE3GroupBase::Tangent = Eigen::Matrix<double, 6, 1>; typename Eigen::internal::traits::Scalar = double]’
/home/panda/Desktop/slam_baselines/dvo_slam/dvo_core/src/dense_tracking.cpp:238:17: required from here
/home/panda/Desktop/slam_baselines/dvo_slam/sophus/include/sophus/so3.hpp:569:3: warning: no return statement in function returning non-void [-Wreturn-type]
}
^
In file included from /home/panda/Desktop/slam_baselines/dvo_slam/dvo_core/src/dense_tracking.cpp:27:0:
/home/panda/Desktop/slam_baselines/dvo_slam/sophus/include/sophus/se3.hpp: In member function ‘Sophus::SE3Group<typename Eigen::internal::traits::Scalar> Sophus::SE3GroupBase::inverse() const [with Derived = Sophus::SE3Group; typename Eigen::internal::traits::Scalar = double]’:
/home/panda/Desktop/slam_baselines/dvo_slam/sophus/include/sophus/se3.hpp:200:3: warning: control reaches end of non-void function [-Wreturn-type]
}
^
/home/panda/Desktop/slam_baselines/dvo_slam/sophus/include/sophus/se3.hpp: In member function ‘Eigen::Matrix<typename Eigen::internal::traits::Scalar, 3, 3> Sophus::SE3GroupBase::rotationMatrix() const [with Derived = Sophus::SE3Group; typename Eigen::internal::traits::Scalar = double]’:
/home/panda/Desktop/slam_baselines/dvo_slam/sophus/include/sophus/se3.hpp:312:3: warning: control reaches end of non-void function [-Wreturn-type]
}
^
In file included from /home/panda/Desktop/slam_baselines/dvo_slam/dvo_core/include/dvo/dense_tracking.h:28:0,
from /home/panda/Desktop/slam_baselines/dvo_slam/dvo_core/src/dense_tracking.cpp:23:
/home/panda/Desktop/slam_baselines/dvo_slam/dvo_core/include/dvo/core/datatypes.h: At global scope:
/home/panda/Desktop/slam_baselines/dvo_slam/dvo_core/include/dvo/core/datatypes.h:33:28: warning: ‘dvo::core::Invalid’ defined but not used [-Wunused-variable]
static const IntensityType Invalid = std::numeric_limits::quiet_NaN();
^
/home/panda/Desktop/slam_baselines/dvo_slam/dvo_core/include/dvo/core/datatypes.h:36:24: warning: ‘dvo::core::InvalidDepth’ defined but not used [-Wunused-variable]
static const DepthType InvalidDepth = std::numeric_limits::quiet_NaN();
^
make[3]: *** [CMakeFiles/dvo_core.dir/src/dense_tracking.o] Error 1
make[3]: Leaving directory /home/panda/Desktop/slam_baselines/dvo_slam/dvo_core/build' make[2]: *** [CMakeFiles/dvo_core.dir/all] Error 2 make[2]: Leaving directory/home/panda/Desktop/slam_baselines/dvo_slam/dvo_core/build'
make[1]: *** [all] Error 2
make[1]: Leaving directory `/home/panda/Desktop/slam_baselines/dvo_slam/dvo_core/build'
-------------------------------------------------------------------------------}
[ rosmake ] Output from build of package dvo_core written to:
[ rosmake ] /home/panda/.ros/rosmake/rosmake_output-20161004-150237/dvo_core/build_output.log
[rosmake-0] Finished <<< dvo_core [FAIL] [ 8.61 seconds ]
[ rosmake ] Halting due to failure in package dvo_core.
[ rosmake ] Waiting for other threads to complete.
[ rosmake ] Results:
[ rosmake ] Built 40 packages with 1 failures.
[ rosmake ] Summary output to directory

Answer 1 · 2016-10-05T02:48:45.000Z

Hi,

A few months ago I attempted to compile DVO_SLAM on Linux Mint (Linux distro based on Ubuntu). The forked repository and step-by-step instructions for compilation are here: https://github.com/songuke/dvo_slam

You might want to give it try. It uses catkin to build, which seems to avoid many issues caused by rosmake.

Cheers.

Answer 2 · 2016-10-10T00:27:10.000Z

Hi,

Thanks for your help.

It doesn't work for me yet. I keep having trouble with type conversion
between Sophus matrix and eigen::affine3f.

Which Eigen version did you use to compile the package? I notice that when
I change the Eigen version, sometimes Sophus cannot be compiled.

Le

On Tue, Oct 4, 2016 at 10:48 PM, songuke notifications@github.com wrote:

Hi,

A few months ago I attempted to compile DVO on Linux Mint (Linux distro
based on Ubuntu). The step-by-step instructions for compilations are here:
https://github.com/songuke/dvo_slam

You might want to give it try. It uses catkin to build, which seems to
avoid many issues caused by rosmake.

Cheers.

—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
#50 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AOXj5OX1Pu8Dd7dihj4n4Gj1K_IZQX3zks5qwxAPgaJpZM4KODT5
.

Answer 3 · 2016-10-10T14:16:44.000Z

Hi Le,

In the README.md on my repository, I did note:

"If your system has Eigen >= 3.3, don't use it with g2o. The old version of
g2o is only compatible to Eigen 3.2.8..."

You might want to try that.

2016-10-10 7:27 GMT+07:00 ple13 notifications@github.com:

Hi,

Thanks for your help.

It doesn't work for me yet. I keep having trouble with type conversion
between Sophus matrix and eigen::affine3f.

Which Eigen version did you use to compile the package? I notice that when
I change the Eigen version, sometimes Sophus cannot be compiled.

Le

On Tue, Oct 4, 2016 at 10:48 PM, songuke notifications@github.com wrote:

Hi,

A few months ago I attempted to compile DVO on Linux Mint (Linux distro
based on Ubuntu). The step-by-step instructions for compilations are
here:
https://github.com/songuke/dvo_slam

You might want to give it try. It uses catkin to build, which seems to
avoid many issues caused by rosmake.

Cheers.

—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#50 (comment)
,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/
AOXj5OX1Pu8Dd7dihj4n4Gj1K_IZQX3zks5qwxAPgaJpZM4KODT5>
.

—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#50 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AAIXslOrgU8TErFq8UeLZmVmh6zl97SFks5qyYZfgaJpZM4KODT5
.

Answer 4 · 2016-10-11T15:24:43.000Z

@ple13 : Have you solved the issue?, I got the identical problem.
@songuke : I already tried your dvo_slam from: https://github.com/songuke/dvo_slam. I've checked my eigen version and it is 3.2.0. Do you have any idea why the Eigen conversion seems not working?

My OS is Ubuntu 14.04 with ROS Indigo.

Answer 5 · 2016-10-11T15:33:53.000Z

I tried to check the installed eigen:

ubuntu@ubuntu:~/dvo_slam$ dpkg --get-selections | grep eigen
libeigen3-dev install
ros-indigo-eigen-conversions install
ros-indigo-eigen-stl-containers install

ros indigo has its own conversion package, will the multiple packages be the possible issue?

Answer 6 · 2016-10-12T20:46:34.000Z

Hi,

Thank you Songuke!!

I finally compiled the code successfully.

I guess that the Eigen type conversion errors that kept popping out was due
to the compiler I used (g++-4.7/4.8.4.9/5). It compiled the Sophus package
without any problems, however, when Sophus was used in dvo_slam, the
compiler complainted. I checked the error messages, then modified the file
so3.hpp in Sophus, and it worked.

In short, I changed these lines in /usr/local/include/Sophus/so3.hpp,
starting from line 89:

/** \brief scalar type /
using Scalar = typename Eigen::internal::traits::Scalar;
/* \brief quaternion reference type /
using QuaternionReference =
typename Eigen::internal::traits::QuaternionType&;
/* \brief quaternion const reference type */
using ConstQuaternionReference =
typename Eigen::internal::traits::QuaternionType const&;

into

/** \brief scalar type /
typedef typename Eigen::internal::traits::Scalar Scalar;
/* \brief quaternion reference type /
typedef
typename Eigen::internal::traits::QuaternionType&
QuaternionReference;
/* \brief quaternion const reference type */
typedef
typename Eigen::internal::traits::QuaternionType const&
ConstQuaternionReference;

After this small change, I could compiled the package with g++.

@mrksst: when I installed Eigen, I used cmake ..
-DCMAKE_INSTALL_PREFIX=/usr . This was to avoid the situation of having
multiple eigens in the system. Before doing that, I deleted all eigen3 in
/usr/include and /usr/local/include, then restalled it using the above
option. I hope that it works for you too.

Le, Phi Hung

On Tue, Oct 11, 2016 at 11:33 AM, mrkstt notifications@github.com wrote:

I tried to check the installed eigen:

ubuntu@ubuntu:~/dvo_slam$ dpkg --get-selections | grep eigen
libeigen3-dev install
ros-indigo-eigen-conversions install
ros-indigo-eigen-stl-containers install

ros indigo has its own conversion package, will the multiple packages be
the possible issue?

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#50 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AOXj5PVH1DtIinUlKHMIdqJi7jtf6dR7ks5qy6xkgaJpZM4KODT5
.

Answer 7 · 2016-10-13T03:35:09.000Z

Hi Le @ple13,
Sounds great you have solved the issue.
I followed the code modification above and the error disappeared. I also re-installed the eigen as you suggested (my Eigen version is 3.2.8). However, I got this following error now:

ubuntu@ubuntu:~/dvo_slam$ catkin_make -DCMAKE_BUILD_TYPE=ReleaseBase path: /home/ubuntu/dvo_slam
Source space: /home/ubuntu/dvo_slam/src
Build space: /home/ubuntu/dvo_slam/build
Devel space: /home/ubuntu/dvo_slam/devel
Install space: /home/ubuntu/dvo_slam/install

Running command: "make cmake_check_build_system" in "/home/ubuntu/dvo_slam/build"

Running command: "make -j8 -l8" in "/home/ubuntu/dvo_slam/build"

[ 7%] [ 7%] Built target dvo_slam_gencfg
Built target dvo_ros_gencfg
[ 10%] Building CXX object dvo_core/CMakeFiles/dvo_core.dir/src/dense_tracking.cpp.o
In file included from /usr/local/include/sophus/sophus.hpp:31:0,
from /usr/local/include/sophus/so3.hpp:27,
from /usr/local/include/sophus/se3.hpp:27,
from /home/ubuntu/dvo_slam/src/dvo_core/src/dense_tracking.cpp:27:
/usr/local/include/sophus/ensure.hpp:60:1: error: ‘EIGEN_DEVICE_FUNC’ does not name a type
EIGEN_DEVICE_FUNC inline void defaultEnsure(const char* function,
^
/usr/local/include/sophus/so3.hpp: In member function ‘void Sophus::SO3GroupBase::normalize()’:
/usr/local/include/sophus/so3.hpp:221:5: error: ‘defaultEnsure’ is not a member of ‘Sophus’
SOPHUS_ENSURE(length >= SophusConstants::epsilon(),
^
/usr/local/include/sophus/so3.hpp: In static member function ‘static Sophus::SO3GroupBase::Transformation Sophus::SO3GroupBase::generator(int)’:
/usr/local/include/sophus/so3.hpp:431:5: error: ‘defaultEnsure’ is not a member of ‘Sophus’
SOPHUS_ENSURE(i >= 0 && i <= 2, "i should be in range [0,2].");
^
/usr/local/include/sophus/so3.hpp: In static member function ‘static void Sophus::SO3GroupBase::internalGenerator(int, Eigen::Quaternion<typename Eigen::internal::traits::Scalar>)’:
/usr/local/include/sophus/so3.hpp:445:5: error: ‘defaultEnsure’ is not a member of ‘Sophus’
SOPHUS_ENSURE(i >= 0 && i <= 2, "i should be in range [0,2]");
^
/usr/local/include/sophus/so3.hpp:446:5: error: ‘defaultEnsure’ is not a member of ‘Sophus’
SOPHUS_ENSURE(internal_gen_q != NULL,
^
/usr/local/include/sophus/so3.hpp: In static member function ‘static Sophus::SO3GroupBase::Tangent Sophus::SO3GroupBase::logAndTheta(const Sophus::SO3Group<typename Eigen::internal::traits::Scalar>&, Sophus::SO3GroupBase::Scalar)’:
/usr/local/include/sophus/so3.hpp:551:7: error: ‘defaultEnsure’ is not a member of ‘Sophus’
SOPHUS_ENSURE(abs(w) >= SophusConstants::epsilon(),
^
/usr/local/include/sophus/se3.hpp: In static member function ‘static Sophus::SE3GroupBase::Transformation Sophus::SE3GroupBase::generator(int)’:
/usr/local/include/sophus/se3.hpp:503:5: error: ‘defaultEnsure’ is not a member of ‘Sophus’
SOPHUS_ENSURE(i >= 0 && i <= 5, "i should be in range [0,5].");
^
/usr/local/include/sophus/se3.hpp: In static member function ‘static void Sophus::SE3GroupBase::internalGenerator(int, Eigen::Quaternion<typename Eigen::internal::traits::Scalar>, Eigen::Matrix<typename Eigen::internal::traits::Scalar, 3, 1>)’:
/usr/local/include/sophus/se3.hpp:519:5: error: ‘defaultEnsure’ is not a member of ‘Sophus’
SOPHUS_ENSURE(i >= 0 && i <= 5, "i should be in range [0,5]");
^
/usr/local/include/sophus/se3.hpp:520:5: error: ‘defaultEnsure’ is not a member of ‘Sophus’
SOPHUS_ENSURE(internal_gen_q != NULL,
^
/usr/local/include/sophus/se3.hpp:522:5: error: ‘defaultEnsure’ is not a member of ‘Sophus’
SOPHUS_ENSURE(internal_gen_t != NULL,
^
make[2]: *** [dvo_core/CMakeFiles/dvo_core.dir/src/dense_tracking.cpp.o] Error 1
make[1]: *** [dvo_core/CMakeFiles/dvo_core.dir/all] Error 2
make: *** [all] Error 2
Invoking "make -j8 -l8" failed

Seems "EIGEN_DEVICE_FUNC" and "defaultEnsure" are not recognized.
Any idea?, really thanks for suggestion.

Answer 8 · 2016-10-13T07:21:26.000Z

Hi @ple13 @songuke,

I solved the error above by commenting all "SOPHUS_ENSURE", seems it's just for checking.

Then, I run:

roslaunch launch/benchmark.launch dataset:=/home/ubuntu/dataset/ rgbd_dataset_freiburg3_cabinet_validation

Output:
ros.roscpp: { Tried to advertise a service that is already advertised in this node [/dvo_vis/image/compressedDepth/set_parameters] }
ros.roscpp: { Tried to advertise a service that is already advertised in this node [/dvo_vis/image/compressed/set_parameters] }
ros.roscpp: { Tried to advertise a service that is already advertised in this node [/dvo_vis/image/theora/set_parameters] }
ros.dvo_benchmark.config: { tracker config: "First Level = 3, Last Level = 1, Max Iterations per Level = 50, Precision = 0.0001, Mu = 0.05, Use Initial Estimate = true, Use Weighting = true, Scale Estimator = TDistribution, Scale Estimator Param = 5, Influence Function = TDistribution, Influence Function Param = 5, Intensity Derivative Threshold = 0, Depth Derivative Threshold = 0" }
ros.dvo_benchmark.config: { frontend config: "UseMultiThreading: 1 MaxTranslationalDistance: 0.2 MaxRotationalDistance: 0 MinEntropyRatio: 0.6 MinEquationSystemConstraintRatio: 0.3" }
ros.dvo_benchmark.config: { backend config: "UseRobustKernel: 1 UseMultiThreading: 1 NewConstraintSearchRadius: 5 NewConstraintMinEntropyRatioCoarse: 0.03 NewConstraintMinEntropyRatioFine: 0.6 MinEquationSystemConstraintRatio: 0.3 MinConstraintDistance: 0 OptimizationUseDenseGraph: 0 OptimizationIterations: 50 OptimizationRemoveOutliers: 1 OptimizationOutlierWeightThreshold: 0.1 FinalOptimizationUseDenseGraph: 1 FinalOptimizationIterations: 1000 FinalOptimizationRemoveOutliers: 1 FinalOptimizationOutlierWeightThreshold: 0.1" }
ros.dvo_slam: { adding 0 new constraints }
ros.dvo_slam: { adding 1 new constraints }
ros.dvo_slam: { adding 2 new constraints }
ros.dvo_slam: { adding 3 new constraints }
ros.dvo_slam: { adding 4 new constraints }
ros.dvo_slam: { adding 5 new constraints }
ros.dvo_slam: { adding 6 new constraints }
ros.dvo_slam: { adding 6 new constraints }
ros.dvo_slam: { adding 6 new constraints }
new_kf: 0.0979787
ros.dvo_slam: { adding 6 new constraints }
ros.dvo_slam: { adding 9 new constraints }
ros.dvo_slam: { adding 11 new constraints }
ros.dvo_slam: { adding 7 new constraints }
ros.dvo_slam: { adding 8 new constraints }
ros.dvo_slam: { adding 9 new constraints }
match: 0.0492692
...

The only thing unsolved is I cannot see any result (empty) by running:

rosrun rviz rviz

Any idea?

Answer 9 · 2016-10-13T08:32:41.000Z

Hi guys,

Thanks for all the suggestions to fix the issues.

My previous experience with rviz is that it does take a while for the
result to appear. Please also make sure the code is compiled in Release
mode.

2016-10-13 14:21 GMT+07:00 mrkstt notifications@github.com:

Hi @ple13 https://github.com/ple13 @songuke https://github.com/songuke
,

I solved the error above by commenting all "SOPHUS_ENSURE", seems it's
just for checking.

Then, I run:

roslaunch launch/benchmark.launch dataset:=/home/ubuntu/dataset/
rgbd_dataset_freiburg3_cabinet_validation

Output:
ros.roscpp: { Tried to advertise a service that is already advertised in
this node [/dvo_vis/image/compressedDepth/set_parameters] }
ros.roscpp: { Tried to advertise a service that is already advertised in
this node [/dvo_vis/image/compressed/set_parameters] }
ros.roscpp: { Tried to advertise a service that is already advertised in
this node [/dvo_vis/image/theora/set_parameters] }
ros.dvo_benchmark.config: { tracker config: "First Level = 3, Last Level =
1, Max Iterations per Level = 50, Precision = 0.0001, Mu = 0.05, Use
Initial Estimate = true, Use Weighting = true, Scale Estimator =
TDistribution, Scale Estimator Param = 5, Influence Function =
TDistribution, Influence Function Param = 5, Intensity Derivative Threshold
= 0, Depth Derivative Threshold = 0" }
ros.dvo_benchmark.config: { frontend config: "UseMultiThreading: 1
MaxTranslationalDistance: 0.2 MaxRotationalDistance: 0 MinEntropyRatio: 0.6
MinEquationSystemConstraintRatio: 0.3" }
ros.dvo_benchmark.config: { backend config: "UseRobustKernel: 1
UseMultiThreading: 1 NewConstraintSearchRadius: 5
NewConstraintMinEntropyRatioCoarse: 0.03 NewConstraintMinEntropyRatioFine:
0.6 MinEquationSystemConstraintRatio: 0.3 MinConstraintDistance: 0
OptimizationUseDenseGraph: 0 OptimizationIterations: 50
OptimizationRemoveOutliers: 1 OptimizationOutlierWeightThreshold: 0.1
FinalOptimizationUseDenseGraph: 1 FinalOptimizationIterations: 1000
FinalOptimizationRemoveOutliers: 1 FinalOptimizationOutlierWeightThreshold:
0.1" }
ros.dvo_slam: { adding 0 new constraints }
ros.dvo_slam: { adding 1 new constraints }
ros.dvo_slam: { adding 2 new constraints }
ros.dvo_slam: { adding 3 new constraints }
ros.dvo_slam: { adding 4 new constraints }
ros.dvo_slam: { adding 5 new constraints }
ros.dvo_slam: { adding 6 new constraints }
ros.dvo_slam: { adding 6 new constraints }
ros.dvo_slam: { adding 6 new constraints }
new_kf: 0.0979787
ros.dvo_slam: { adding 6 new constraints }
ros.dvo_slam: { adding 9 new constraints }
ros.dvo_slam: { adding 11 new constraints }
ros.dvo_slam: { adding 7 new constraints }
ros.dvo_slam: { adding 8 new constraints }
ros.dvo_slam: { adding 9 new constraints }
match: 0.0492692

The only thing unsolved is I cannot see any result (empty) by running:

rosrun rviz rviz

Any idea?

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#50 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AAIXsgpqjdzA8UDaf_hP89eCUdAOMW5uks5qzdv3gaJpZM4KODT5
.

Answer 10 · 2016-10-13T18:20:39.000Z

Hi,

When I tried to run the program, it kept asking for the file assoc.txt
however this file is not provided with the dataset. Am I suppose to sync
the depth and rgb to produce this file myself?

Le, Phi Hung

On Thu, Oct 13, 2016 at 4:32 AM, songuke notifications@github.com wrote:

Hi guys,

Thanks for all the suggestions to fix the issues.

My previous experience with rviz is that it does take a while for the
result to appear. Please also make sure the code is compiled in Release
mode.

2016-10-13 14:21 GMT+07:00 mrkstt notifications@github.com:

Hi @ple13 https://github.com/ple13 @songuke <
https://github.com/songuke>
,

I solved the error above by commenting all "SOPHUS_ENSURE", seems it's
just for checking.

Then, I run:

roslaunch launch/benchmark.launch dataset:=/home/ubuntu/dataset/
rgbd_dataset_freiburg3_cabinet_validation

Output:
ros.roscpp: { Tried to advertise a service that is already advertised in
this node [/dvo_vis/image/compressedDepth/set_parameters] }
ros.roscpp: { Tried to advertise a service that is already advertised in
this node [/dvo_vis/image/compressed/set_parameters] }
ros.roscpp: { Tried to advertise a service that is already advertised in
this node [/dvo_vis/image/theora/set_parameters] }

ros.dvo_benchmark.config: { tracker config: "First Level = 3, Last Level

1, Max Iterations per Level = 50, Precision = 0.0001, Mu = 0.05, Use
Initial Estimate = true, Use Weighting = true, Scale Estimator =
TDistribution, Scale Estimator Param = 5, Influence Function =
TDistribution, Influence Function Param = 5, Intensity Derivative
Threshold
= 0, Depth Derivative Threshold = 0" }
ros.dvo_benchmark.config: { frontend config: "UseMultiThreading: 1
MaxTranslationalDistance: 0.2 MaxRotationalDistance: 0 MinEntropyRatio:
0.6
MinEquationSystemConstraintRatio: 0.3" }
ros.dvo_benchmark.config: { backend config: "UseRobustKernel: 1
UseMultiThreading: 1 NewConstraintSearchRadius: 5
NewConstraintMinEntropyRatioCoarse: 0.03 NewConstraintMinEntropyRatioFi
ne:
0.6 MinEquationSystemConstraintRatio: 0.3 MinConstraintDistance: 0
OptimizationUseDenseGraph: 0 OptimizationIterations: 50
OptimizationRemoveOutliers: 1 OptimizationOutlierWeightThreshold: 0.1
FinalOptimizationUseDenseGraph: 1 FinalOptimizationIterations: 1000
FinalOptimizationRemoveOutliers: 1 FinalOptimizationOutlierWeight
Threshold:
0.1" }
ros.dvo_slam: { adding 0 new constraints }
ros.dvo_slam: { adding 1 new constraints }
ros.dvo_slam: { adding 2 new constraints }
ros.dvo_slam: { adding 3 new constraints }
ros.dvo_slam: { adding 4 new constraints }
ros.dvo_slam: { adding 5 new constraints }
ros.dvo_slam: { adding 6 new constraints }
ros.dvo_slam: { adding 6 new constraints }
ros.dvo_slam: { adding 6 new constraints }
new_kf: 0.0979787
ros.dvo_slam: { adding 6 new constraints }
ros.dvo_slam: { adding 9 new constraints }
ros.dvo_slam: { adding 11 new constraints }
ros.dvo_slam: { adding 7 new constraints }
ros.dvo_slam: { adding 8 new constraints }
ros.dvo_slam: { adding 9 new constraints }
match: 0.0492692

The only thing unsolved is I cannot see any result (empty) by running:

rosrun rviz rviz

Any idea?

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#50 (comment)
,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AAIXsgpqjdzA8UDaf_
hP89eCUdAOMW5uks5qzdv3gaJpZM4KODT5>
.

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#50 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AOXj5P_GYvG6-z7WcbnvpokYLFIaSfFuks5qzeysgaJpZM4KODT5
.

Answer 11 · 2016-10-13T19:17:52.000Z

assoc.txt can be generated with this Python script:
http://vision.in.tum.de/data/datasets/rgbd-dataset/tools

Another hint to check if the program is working is to monitor
trajectory.txt. This file should be changing if DVO is doing the
calculation.

2016-10-14 1:20 GMT+07:00 ple13 notifications@github.com:

Hi,

When I tried to run the program, it kept asking for the file assoc.txt
however this file is not provided with the dataset. Am I suppose to sync
the depth and rgb to produce this file myself?

Le, Phi Hung

On Thu, Oct 13, 2016 at 4:32 AM, songuke notifications@github.com wrote:

Hi guys,

Thanks for all the suggestions to fix the issues.

My previous experience with rviz is that it does take a while for the
result to appear. Please also make sure the code is compiled in Release
mode.

2016-10-13 14:21 GMT+07:00 mrkstt notifications@github.com:

Hi @ple13 https://github.com/ple13 @songuke <
https://github.com/songuke>
,

I solved the error above by commenting all "SOPHUS_ENSURE", seems it's
just for checking.

Then, I run:

roslaunch launch/benchmark.launch dataset:=/home/ubuntu/dataset/
rgbd_dataset_freiburg3_cabinet_validation

Output:
ros.roscpp: { Tried to advertise a service that is already advertised
in
this node [/dvo_vis/image/compressedDepth/set_parameters] }
ros.roscpp: { Tried to advertise a service that is already advertised
in
this node [/dvo_vis/image/compressed/set_parameters] }
ros.roscpp: { Tried to advertise a service that is already advertised
in
this node [/dvo_vis/image/theora/set_parameters] }
ros.dvo_benchmark.config: { tracker config: "First Level = 3, Last

Level

1, Max Iterations per Level = 50, Precision = 0.0001, Mu = 0.05, Use
Initial Estimate = true, Use Weighting = true, Scale Estimator =
TDistribution, Scale Estimator Param = 5, Influence Function =
TDistribution, Influence Function Param = 5, Intensity Derivative
Threshold
= 0, Depth Derivative Threshold = 0" }
ros.dvo_benchmark.config: { frontend config: "UseMultiThreading: 1
MaxTranslationalDistance: 0.2 MaxRotationalDistance: 0 MinEntropyRatio:
0.6
MinEquationSystemConstraintRatio: 0.3" }
ros.dvo_benchmark.config: { backend config: "UseRobustKernel: 1
UseMultiThreading: 1 NewConstraintSearchRadius: 5
NewConstraintMinEntropyRatioCoarse: 0.03
NewConstraintMinEntropyRatioFi
ne:
0.6 MinEquationSystemConstraintRatio: 0.3 MinConstraintDistance: 0
OptimizationUseDenseGraph: 0 OptimizationIterations: 50
OptimizationRemoveOutliers: 1 OptimizationOutlierWeightThreshold: 0.1
FinalOptimizationUseDenseGraph: 1 FinalOptimizationIterations: 1000
FinalOptimizationRemoveOutliers: 1 FinalOptimizationOutlierWeight
Threshold:
0.1" }
ros.dvo_slam: { adding 0 new constraints }
ros.dvo_slam: { adding 1 new constraints }
ros.dvo_slam: { adding 2 new constraints }
ros.dvo_slam: { adding 3 new constraints }
ros.dvo_slam: { adding 4 new constraints }
ros.dvo_slam: { adding 5 new constraints }
ros.dvo_slam: { adding 6 new constraints }
ros.dvo_slam: { adding 6 new constraints }
ros.dvo_slam: { adding 6 new constraints }
new_kf: 0.0979787
ros.dvo_slam: { adding 6 new constraints }
ros.dvo_slam: { adding 9 new constraints }
ros.dvo_slam: { adding 11 new constraints }
ros.dvo_slam: { adding 7 new constraints }
ros.dvo_slam: { adding 8 new constraints }
ros.dvo_slam: { adding 9 new constraints }
match: 0.0492692

The only thing unsolved is I cannot see any result (empty) by running:

rosrun rviz rviz

Any idea?

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#50
issuecomment-253434903
,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AAIXsgpqjdzA8UDaf_
hP89eCUdAOMW5uks5qzdv3gaJpZM4KODT5>
.

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#50 (comment)
,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AOXj5P_GYvG6-
z7WcbnvpokYLFIaSfFuks5qzeysgaJpZM4KODT5>
.

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#50 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AAIXsrSX5wrEEnqVnug7u_YzEjb-HWrGks5qznZ4gaJpZM4KODT5
.

Answer 12 · 2016-10-13T21:19:19.000Z

Thanks.

It works for me.

Le, Phi Hung

On Thu, Oct 13, 2016 at 3:17 PM, songuke notifications@github.com wrote:

assoc.txt can be generated with this Python script:
http://vision.in.tum.de/data/datasets/rgbd-dataset/tools

Another hint to check if the program is working is to monitor
trajectory.txt. This file should be changing if DVO is doing the
calculation.

2016-10-14 1:20 GMT+07:00 ple13 notifications@github.com:

Hi,

When I tried to run the program, it kept asking for the file assoc.txt
however this file is not provided with the dataset. Am I suppose to sync
the depth and rgb to produce this file myself?

Le, Phi Hung

On Thu, Oct 13, 2016 at 4:32 AM, songuke notifications@github.com
wrote:

Hi guys,

Thanks for all the suggestions to fix the issues.

My previous experience with rviz is that it does take a while for the
result to appear. Please also make sure the code is compiled in Release
mode.

2016-10-13 14:21 GMT+07:00 mrkstt notifications@github.com:

Hi @ple13 https://github.com/ple13 @songuke <
https://github.com/songuke>
,

I solved the error above by commenting all "SOPHUS_ENSURE", seems
it's
just for checking.

Then, I run:

roslaunch launch/benchmark.launch dataset:=/home/ubuntu/dataset/
rgbd_dataset_freiburg3_cabinet_validation

Output:
ros.roscpp: { Tried to advertise a service that is already advertised
in
this node [/dvo_vis/image/compressedDepth/set_parameters] }
ros.roscpp: { Tried to advertise a service that is already advertised
in
this node [/dvo_vis/image/compressed/set_parameters] }
ros.roscpp: { Tried to advertise a service that is already advertised
in
this node [/dvo_vis/image/theora/set_parameters] }
ros.dvo_benchmark.config: { tracker config: "First Level = 3, Last

Level

1, Max Iterations per Level = 50, Precision = 0.0001, Mu = 0.05, Use
Initial Estimate = true, Use Weighting = true, Scale Estimator =
TDistribution, Scale Estimator Param = 5, Influence Function =
TDistribution, Influence Function Param = 5, Intensity Derivative
Threshold
= 0, Depth Derivative Threshold = 0" }
ros.dvo_benchmark.config: { frontend config: "UseMultiThreading: 1
MaxTranslationalDistance: 0.2 MaxRotationalDistance: 0
MinEntropyRatio:
0.6
MinEquationSystemConstraintRatio: 0.3" }
ros.dvo_benchmark.config: { backend config: "UseRobustKernel: 1
UseMultiThreading: 1 NewConstraintSearchRadius: 5
NewConstraintMinEntropyRatioCoarse: 0.03
NewConstraintMinEntropyRatioFi
ne:
0.6 MinEquationSystemConstraintRatio: 0.3 MinConstraintDistance: 0
OptimizationUseDenseGraph: 0 OptimizationIterations: 50
OptimizationRemoveOutliers: 1 OptimizationOutlierWeightThreshold:
0.1
FinalOptimizationUseDenseGraph: 1 FinalOptimizationIterations: 1000
FinalOptimizationRemoveOutliers: 1 FinalOptimizationOutlierWeight
Threshold:
0.1" }
ros.dvo_slam: { adding 0 new constraints }
ros.dvo_slam: { adding 1 new constraints }
ros.dvo_slam: { adding 2 new constraints }
ros.dvo_slam: { adding 3 new constraints }
ros.dvo_slam: { adding 4 new constraints }
ros.dvo_slam: { adding 5 new constraints }
ros.dvo_slam: { adding 6 new constraints }
ros.dvo_slam: { adding 6 new constraints }
ros.dvo_slam: { adding 6 new constraints }
new_kf: 0.0979787
ros.dvo_slam: { adding 6 new constraints }
ros.dvo_slam: { adding 9 new constraints }
ros.dvo_slam: { adding 11 new constraints }
ros.dvo_slam: { adding 7 new constraints }
ros.dvo_slam: { adding 8 new constraints }
ros.dvo_slam: { adding 9 new constraints }
match: 0.0492692

The only thing unsolved is I cannot see any result (empty) by
running:

rosrun rviz rviz

Any idea?

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#50
issuecomment-253434903
,
or mute the thread
<https://github.com/notifications/unsubscribe-
auth/AAIXsgpqjdzA8UDaf_
hP89eCUdAOMW5uks5qzdv3gaJpZM4KODT5>
.

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#50
issuecomment-253449258
,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AOXj5P_GYvG6-
z7WcbnvpokYLFIaSfFuks5qzeysgaJpZM4KODT5>
.

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#50 (comment)
,
or mute the thread
<https://github.com/notifications/unsubscribe-
auth/AAIXsrSX5wrEEnqVnug7u_YzEjb-HWrGks5qznZ4gaJpZM4KODT5>

.

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#50 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AOXj5BiN-1Y4SIRG94eR2pQ7Nendtk1Pks5qzoPjgaJpZM4KODT5
.