petititi / gsoc2011sfm Goto Github PK
View Code? Open in Web Editor NEWAutomatically exported from code.google.com/p/gsoc2011sfm
Automatically exported from code.google.com/p/gsoc2011sfm
What steps will reproduce the problem?
1. checkout r210
2. cmake && make
What is the expected output? What do you see instead?
Compile fails due to absence of #DEFINEs located in config.h
Copying <gsoc_root>/config.h.in to src/ and tutorials/ solves the problem.
What version of the product are you using? On what operating system?
r210
Ubuntu 11.04
Please provide any additional information below.
Not 100% sure this is your issue, thibauld. I'm not very familiar with cmake.
Original issue reported on code.google.com by [email protected]
on 10 Aug 2011 at 9:51
What steps will reproduce the problem?
1. Complete all make/install steps.
2. Attempt to compile library.
What is the expected output? What do you see instead?
A handful of compile-time errors occurred relating to libmv.
I'm not sure whether these are due to a previous configuration/installation
error, but I think it's still useful to document. It seems that the code is
able to accept declarations in the libmv namespace but not call functions -
maybe a linker problem?
In the meantime, I have been able to comment out the code which caused these
errors and continue to use the other parts of the library.
Please use labels and text to provide additional information.
== ERROR 1 ==
/home/steve/OpenCV/gsoc2011sfm/src/EuclideanEstimator.cpp||In function
‘double OpencvSfM::SampsonDistance2(const libmv::Mat&, const libmv::Mat2X&,
const libmv::Mat2X&)’:|
/home/steve/OpenCV/gsoc2011sfm/src/EuclideanEstimator.cpp|34|error: ‘SampsonDistance’ is not a member of ‘libmv’|
== ERROR 2 ==
/home/steve/OpenCV/gsoc2011sfm/src/EuclideanEstimator.cpp||In function ‘double OpencvSfM::robust5Points(const libmv::Mat2X&, const libmv::Mat2X&, const libmv::Mat3&, const libmv::Mat3&, libmv::Mat3&)’:|
/home/steve/OpenCV/gsoc2011sfm/src/EuclideanEstimator.cpp|385|error: ‘robust5Points’ was not declared in this scope|
== ERROR 3 ==
../src/libSfM_core.so.0.1.0||undefined reference to
`libmv::MotionFromEssentialAndCorrespondence(Eigen::Matrix<double, 3, 3, 0, 3,
3> const&, Eigen::Matrix<double, 3, 3, 0, 3, 3> const&, Eigen::Matrix<double,
2, 1, 0, 2, 1> const&, Eigen::Matrix<double, 3, 3, 0, 3, 3> const&,
Eigen::Matrix<double, 2, 1, 0, 2, 1> const&, Eigen::Matrix<double, 3, 3, 0, 3,
3>*, Eigen::Matrix<double, 3, 1, 0, 3, 1>*)'|
../src/libSfM_core.so.0.1.0||undefined reference to
`libmv::EuclideanResectionEPnPRobust(Eigen::Matrix<double, 2, -1, 0, 2, -1>
const&, Eigen::Matrix<double, 3, -1, 0, 3, -1> const&, Eigen::Matrix<double, 3,
3, 0, 3, 3> const&, double, Eigen::Matrix<double, 3, 3, 0, 3, 3>*,
Eigen::Matrix<double, 3, 1, 0, 3, 1>*, libmv::vector<int,
Eigen::aligned_allocator<int> >*, double)'|
../src/libSfM_core.so.0.1.0||undefined reference to
`libmv::EuclideanBAFull(libmv::vector<Eigen::Matrix<double, 2, -1, 0, 2, -1>,
Eigen::aligned_allocator<Eigen::Matrix<double, 2, -1, 0, 2, -1> > > const&,
libmv::vector<Eigen::Matrix<double, 3, 3, 0, 3, 3>,
Eigen::aligned_allocator<Eigen::Matrix<double, 3, 3, 0, 3, 3> > >*,
libmv::vector<Eigen::Matrix<double, 3, 3, 0, 3, 3>,
Eigen::aligned_allocator<Eigen::Matrix<double, 3, 3, 0, 3, 3> > >*,
libmv::vector<Eigen::Matrix<double, 3, 1, 0, 3, 1>,
Eigen::aligned_allocator<Eigen::Matrix<double, 3, 1, 0, 3, 1> > >*,
Eigen::Matrix<double, 3, -1, 0, 3, -1>*, libmv::eLibmvBundleType)'|
Original issue reported on code.google.com by [email protected]
on 1 Aug 2011 at 3:14
What steps will reproduce the problem?
1. Run Tutorial #4
What is the expected output? What do you see instead?
It should display a fairly dense point cloud. Instead, number of correct tracks
loaded appears to be 0 and no model is displayed.
Please use labels and text to provide additional information.
The function motion_estim_loaded->keepOnlyCorrectMatches() is called in
tutorials/test_PCL_visualization.cpp. The default arguments (min_matches = 20,
min_consistance = 3) are too harsh and result in the rejection of ALL tracks.
If you don't want to change the defaults, I'd suggest passing some arguments
into "keepOnlyCorrectMatches()" in order to preserve a higher proportion of
tracks.
Additionally, I think there is an error in
SequenceAnalyzer::keepOnlyCorrectMatches() function. Although minimum values
are defined, points are rejected even if they are equal to these minimums. I'd
suggest changing the line:
if( ( tracks_[ index ].getNbTrack( ) <= min_matches ) ||
( tracks_[ index ].track_consistance <= (int)min_consistance ) )
to
if( ( tracks_[ index ].getNbTrack( ) < min_matches ) ||
( tracks_[ index ].track_consistance < (int)min_consistance ) )
To make it more intuitive.
Please let me know if you want me to make these changes and I will commit and
remove this bug, but I thought it would be best to post it here first for
approval.
Original issue reported on code.google.com by [email protected]
on 19 Aug 2011 at 4:38
What steps will reproduce the problem?
1. Run Tutorial #9 and select option 1
What is the expected output? What do you see instead?
It should load data and then show a 3D point cloud with camera positions.
Instead it seg faults.
Please use labels and text to provide additional information.
The line causing the seg fault is in src/Visualizer.cpp:
pcl::io::saveVTKFile ( ((std::string)"test")+name+((std::string)".vtk"),
triangles );
Was this just in there for testing purposes? When commented out the tutorial
runs fine, and the tutorial doesn't imply that it will write out any data
anyway. As always, just give me the go ahead if you want me to commit.
Original issue reported on code.google.com by [email protected]
on 19 Aug 2011 at 5:06
Purpose of code changes on this branch:
Avoiding a compile-time error relating to the eigen libraries.
When reviewing my code changes, please focus on:
Just make sure it doesn't break things on your machine. On mine it successfully
avoided the following compile-time error:
/usr/include/eigen3/Eigen/Array|8|error: #error The Eigen/Array header does no
longer exist in Eigen3. All that functionality has moved to Eigen/Core.|
After the review, I'll merge this branch into:
/trunk
Original issue reported on code.google.com by [email protected]
on 31 Jul 2011 at 4:09
Attachments:
What steps will reproduce the problem?
1. Check out source from repository.
2. Open CMake and attempt to configure.
What is the expected output? What do you see instead?
Ideally the configuration should complete without errors. Instead, an error
occurs because it cannot find LIBMV_HEADER_DIR, although it does find
LIBMV_LIB_DIR correctly. I think I was able to manually correct this in Cmake
by putting in the address: /libmv/libmv-0.1.1104-dev/src, but I'm not sure how
to get CMake to locate this automatically..
Please use labels and text to provide additional information.
Looking for pthread_create in pthread
Looking for pthread_create in pthread - found
Found Threads: TRUE
Found wxWidgets: TRUE
CMake Error: The following variables are used in this project, but they are set
to NOTFOUND.
Please set them or make sure they are set and tested correctly in the CMake
files:
LIBMV_HEADER_DIR
used as include directory in directory /home/steve/OpenCV/gsoc2011sfm
used as include directory in directory /home/steve/OpenCV/gsoc2011sfm/src
used as include directory in directory /home/steve/OpenCV/gsoc2011sfm/tutorials
Configuring incomplete, errors occurred!
Original issue reported on code.google.com by [email protected]
on 31 Jul 2011 at 4:01
What steps will reproduce the problem?
1. Run Tutorial #9 and select option 0
What is the expected output? What do you see instead?
Expected: Similar results to Tutorial #9 option 1 (once it is working).
Basically should show an accurate point cloud and camera positions, but all
cameras are stacked on top of each other.
Please use labels and text to provide additional information.
This I assume is the same bundle adjustment problem that you mentioned on the
wiki, which you said you fixed, right? Maybe the tutorial just needs to be
updated to use the corrected method.
Original issue reported on code.google.com by [email protected]
on 19 Aug 2011 at 5:12
What steps will reproduce the problem?
1. Fresh checkout, cmake and build (maybe just on linux..)
What is the expected output? What do you see instead?
Should build fine. Instead:
/home/steve/OpenCV/gsoc2011sfm/YAML_loader/main.cpp|144|error: no matching
function for call to
‘transform(
__gnu_cxx::__normal_iterator<char*, std::basic_string<char,
std::char_traits<char>, std::allocator<char> > >,
__gnu_cxx::__normal_iterator<char*, std::basic_string<char,
std::char_traits<char>, std::allocator<char> > >,
__gnu_cxx::__normal_iterator<char*, std::basic_string<char,
std::char_traits<char>, std::allocator<char> > >,
<unresolved overloaded function type>)’|
Please use labels and text to provide additional information.
After switching to eigen-3.0.1 like you suggested it compiled and ran fine
(although a couple of tests still weren't working). I then tried a clean
checkout and build and suddenly I am getting this error when I try to
compile... :/
Maybe it has something to do with this warning in CMake:
CMake Warning at ProjectAdd.cmake:87 (add_executable):
Cannot generate a safe linker search path for target SfM_YAML_loader
because files in some directories may conflict with libraries in implicit
directories:
link library [libpng.so] in /usr/lib may be hidden by files in:
/usr/local/share/opencv/../../lib
/usr/local/lib
Some of these libraries may not be found correctly.
Call Stack (most recent call first):
YAML_loader/CMakeLists.txt:2 (new_executable)
Original issue reported on code.google.com by [email protected]
on 5 Sep 2011 at 12:34
while compiling the project in
http://gsoc2011sfm.googlecode.com/svntrunk/trunk/, i got the error message:
"C:\Program Files\PCL
1.4.0\3rdParty\Boost\include\boost/iterator/iterator_facade.hpp(485) : error
C2059: syntax error : 'string'"
"C:\Program Files\PCL
1.4.0\3rdParty\Boost\include\boost/iterator/iterator_facade.hpp(485) : error
C2143: syntax error : missing ')' before '.'"
I can not realize what the issue is, please help me to solve out.
I use CMAKE 2.8 to make the project for Microsoft Visual Studio 2008 in Windows
7 Ultimate 32 bit.
Thanks a lot.
Original issue reported on code.google.com by [email protected]
on 5 Mar 2012 at 4:23
What steps will reproduce the problem?
1. Check out latest code
2. Configure/Generate in CMake
3. Attempt to compile (I used Codeblocks on Ubuntu 10.10)
What is the expected output? What do you see instead?
It should compile well, as it did before. However, a whole heap of errors
relating to the following file occur:
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h
Please use labels and text to provide additional information.
FULL OUTPUT:
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h||In
constructor ‘Eigen::eigen2_AlignedBox<_Scalar,
_AmbientDim>::eigen2_AlignedBox(int)’:|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h|54|error:
class ‘Eigen::eigen2_AlignedBox<_Scalar, _AmbientDim>’ does not have any
field named ‘m_’|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h|54|error:
expected ‘{’ before ‘(’ token|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h||In
constructor ‘Eigen::eigen2_AlignedBox<_Scalar,
_AmbientDim>::eigen2_AlignedBox(const Eigen::Matrix<_Scalar,
AmbientDimAtCompileTime, 1, ((Eigen::._170)0u | (((AmbientDimAtCompileTime ==
1) && (1 != 1)) ? (Eigen::._170)1u : (((1 == 1) && (AmbientDimAtCompileTime !=
1)) ? (Eigen::._170)0u : (Eigen::._170)0u))), AmbientDimAtCompileTime, 1>&,
const Eigen::Matrix<_Scalar, AmbientDimAtCompileTime, 1, ((Eigen::._170)0u |
(((AmbientDimAtCompileTime == 1) && (1 != 1)) ? (Eigen::._170)1u : (((1 == 1)
&& (AmbientDimAtCompileTime != 1)) ? (Eigen::._170)0u : (Eigen::._170)0u))),
AmbientDimAtCompileTime, 1>&)’:|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h|58|error:
class ‘Eigen::eigen2_AlignedBox<_Scalar, _AmbientDim>’ does not have any
field named ‘m_’|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h|58|error:
expected ‘{’ before ‘(’ token|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h||In
constructor ‘Eigen::eigen2_AlignedBox<_Scalar,
_AmbientDim>::eigen2_AlignedBox(const Eigen::Matrix<_Scalar,
AmbientDimAtCompileTime, 1, ((Eigen::._170)0u | (((AmbientDimAtCompileTime ==
1) && (1 != 1)) ? (Eigen::._170)1u : (((1 == 1) && (AmbientDimAtCompileTime !=
1)) ? (Eigen::._170)0u : (Eigen::._170)0u))), AmbientDimAtCompileTime, 1>&)’:|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h|61|error:
class ‘Eigen::eigen2_AlignedBox<_Scalar, _AmbientDim>’ does not have any
field named ‘m_’|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h|61|error:
expected ‘{’ before ‘(’ token|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h||In member
function ‘void Eigen::eigen2_AlignedBox<_Scalar, _AmbientDim>::setNull()’:|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h|74|error:
expected unqualified-id before ‘(’ token|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h|75|error:
expected unqualified-id before ‘(’ token|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h||In member
function ‘bool Eigen::eigen2_AlignedBox<_Scalar, _AmbientDim>::contains(const
Eigen::eigen2_AlignedBox<_Scalar, _AmbientDim>&) const’:|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h|93|error:
expected unqualified-id before ‘(’ token|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h|93|error:
expected unqualified-id before ‘(’ token|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h||In member
function ‘Eigen::eigen2_AlignedBox<_Scalar, _AmbientDim>&
Eigen::eigen2_AlignedBox<_Scalar, _AmbientDim>::extend(const
Eigen::Matrix<_Scalar, AmbientDimAtCompileTime, 1, ((Eigen::._170)0u |
(((AmbientDimAtCompileTime == 1) && (1 != 1)) ? (Eigen::._170)1u : (((1 == 1)
&& (AmbientDimAtCompileTime != 1)) ? (Eigen::._170)0u : (Eigen::._170)0u))),
AmbientDimAtCompileTime, 1>&)’:|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h|97|error:
expected unqualified-id before ‘(’ token|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h|97|error:
expected unqualified-id before ‘(’ token|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h||In member
function ‘Eigen::eigen2_AlignedBox<_Scalar, _AmbientDim>&
Eigen::eigen2_AlignedBox<_Scalar, _AmbientDim>::extend(const
Eigen::eigen2_AlignedBox<_Scalar, _AmbientDim>&)’:|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h|101|error:
expected unqualified-id before ‘(’ token|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h|101|error:
expected unqualified-id before ‘(’ token|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h||In member
function ‘Eigen::eigen2_AlignedBox<_Scalar, _AmbientDim>&
Eigen::eigen2_AlignedBox<_Scalar, _AmbientDim>::clamp(const
Eigen::eigen2_AlignedBox<_Scalar, _AmbientDim>&)’:|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h|105|error:
expected unqualified-id before ‘(’ token|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h|105|error:
expected unqualified-id before ‘(’ token|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h||In
constructor ‘Eigen::eigen2_AlignedBox<_Scalar,
_AmbientDim>::eigen2_AlignedBox(const Eigen::eigen2_AlignedBox<NewScalarType,
AmbientDimAtCompileTime>&)’:|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h|141|error:
expected unqualified-id before ‘(’ token|
/usr/include/eigen3/Eigen/src/Eigen2Support/Geometry/AlignedBox.h|142|error:
expected unqualified-id before ‘(’ token|
||=== Build finished: 18 errors, 0 warnings ===|
Original issue reported on code.google.com by [email protected]
on 1 Sep 2011 at 8:08
This method don't work, probably due to a wrong method.
When time or if this method is really needed, I will dig deeper, but for now
other methods are more important...
Original issue reported on code.google.com by [email protected]
on 31 May 2011 at 4:12
For now MotionProcessor.h is based on the old "dirent.h" header which is POSIX
library (windows doesn't have this) AND C coding style.
Vincent proposed to use Boost::Filesystem for better file handling.
Original issue reported on code.google.com by [email protected]
on 2 Jun 2011 at 9:20
What steps will reproduce the problem?
1. Run Tutorial #11
What is the expected output? What do you see instead?
Expected: Good SFM result with camera positions.
Actual: Sometimes this is the case, but often not. The result seems to be
random, even though the data is always the same, right?
Please use labels and text to provide additional information.
I've tried using both the LOAD_FULL and LOAD_INTRA flags with the
loadCamerasFromFile() function, and with both strategies the results seem to be
random.
Original issue reported on code.google.com by [email protected]
on 19 Aug 2011 at 5:18
What steps will reproduce the problem?
1. checkout r210
2. cmake-gui && make
What is the expected output? What do you see instead?
linking fails due to undefined refs contained in -lglog -lgflags, which are
required by libmv libs used in tutorials
Adding "-lglog -lgflags" to SfMTutorials.dir/link.txt solves it
What version of the product are you using? On what operating system?
r210
Ubuntu 11.04 64bit
Original issue reported on code.google.com by [email protected]
on 10 Aug 2011 at 10:24
The steps is: running EuclideanReconstruction.
Following code produce error:
SequenceAnalyzer motion_estim( vec_point_for_track, &images, matcher );
cout << "Compute Matches " << endl;
motion_estim.computeMatches( ); //ERROR rising in this method
PointsToTrackWithImage::impl_computeDescriptors_( )
{
descriptor_detector_->compute(
imageToAnalyse_,this->keypoints_,this->descriptors_ ); //CRASH
}
What is the expected output?
Reconstruction
What do you see instead?
Program crashing
What version of the product are you using?
trunc from 30.10.2011
On what operating system?
Windows XP SP2
Please provide any additional information below.
IDE: Eclipse helios 6.5, compiller: MinGW (gcc 4.6.0), lang: Russian
Original issue reported on code.google.com by [email protected]
on 30 Oct 2011 at 12:56
Attachments:
On make the library try to write on /usr/local/
Original issue reported on code.google.com by [email protected]
on 9 Sep 2011 at 1:04
What steps will reproduce the problem?
1. Run Tutorial #0
What is the expected output? What do you see instead?
It should be able to write the results to a file, before reading them back in.
Please use labels and text to provide additional information.
Maybe the new YAML loader hasn't been integrated into the tutorial yet? Because
currently the function "PointsToTrack::write( cv::FileStorage& fs, const
PointsToTrack& keypoints )" is failing..
The line fs << "{" << "PointsToTrack" << "{"; seems to be the first source of
the problem. I think this sort of syntax would work with normal fstream, but
not with the OpenCV "FileStorage" class.
Original issue reported on code.google.com by [email protected]
on 19 Aug 2011 at 4:18
What steps will reproduce the problem?
1. Compile and run EuclideanReconstruction
What is the expected output?
Reconstrucntion
What do you see instead?
Unordered points
What version of the product are you using?
october 2011
On what operating system?
WinXP on VirtualBox
Please provide any additional information below.
first six pictures - Images;
last - Reconstruction result
Original issue reported on code.google.com by [email protected]
on 3 Nov 2011 at 11:25
Attachments:
As this inverse transformation is not trivial, I prefer to wait Vincent's
advises to start the implementation.
Original issue reported on code.google.com by [email protected]
on 31 May 2011 at 4:15
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.