Git Product home page Git Product logo

Comments (11)

HuangTY96 avatar HuangTY96 commented on June 28, 2024

i use ubuntu16.04, and I have installed all the dependency. After that I can run cmake successfully, and run make -j (without badslam) successfully, but I cannot get executable file.

from badslam.

puzzlepaint avatar puzzlepaint commented on June 28, 2024

Hi, do you have the CUDA SDK installed (and does the CMake output indicate that it is found by CMake)? The badslam target is only added if CMAKE_CUDA_COMPILER is set.

from badslam.

HuangTY96 avatar HuangTY96 commented on June 28, 2024

I build the samples file in /usr/local/cuda-9.1 and delete the file build_RelWithDebInfo, then rebuild all the project, finally it works.

from badslam.

HuangTY96 avatar HuangTY96 commented on June 28, 2024

well, when I run ./badslam and choose dataset, then click start, I get this error:
./badslam: symbol lookup error: ./badslam: undefined symbol: _ZN5DBoW26FBrief10fromStringERN5boost14dynamic_bitsetImSaImEEERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEE
could you please tell me what it is and how to solve this

from badslam.

HuangTY96 avatar HuangTY96 commented on June 28, 2024

My dear friend, now when I click start, It can work, but the result is totally horrible, and it gives me an error:
15:01:34.848 loop_detector.cc:355 INFO| - loop closure 3D-3D inlier count: 28
15:01:34.870 loop_detector.cc:601 INFO| - rotational_distance (0, 1): 0.000690534
15:01:34.870 loop_detector.cc:602 INFO| - translational_distance (0, 1): 0.000822317
15:01:34.870 loop_detector.cc:601 INFO| - rotational_distance (0, 2): 0.00231612
15:01:34.870 loop_detector.cc:602 INFO| - translational_distance (0, 2): 0.0025318
15:01:34.870 loop_detector.cc:601 INFO| - rotational_distance (1, 2): 0.00284715
15:01:34.870 loop_detector.cc:602 INFO| - translational_distance (1, 2): 0.00299204
15:01:34.870 loop_detector.cc:668 INFO| - average_pixel_distance: -nan
15:01:34.870 pose_graph_optimizer.cc:127 INFO| - Performing pose graph optimization ...
int g2o::csparse_extension::cs_cholsolsymb(const cs_di*, number_t*, const cs_dis*, number_t*, int*): cholesky failed!
Cholesky failure, writing debug.txt (Hessian loadable by Octave)
15:01:34.918 pose_graph_optimizer.cc:131 INFO| - Pose graph optimization done
Sophus ensure failed in function 'void Sophus::SO3Base::normalize() [with Derived = Sophus::SO3]', file '/home/hty/workspace/badslam/libvis/third_party/sophus/sophus/so3.hpp', line 166.
Quaternion should not be close to zero!
Aborted (core dumped)

this is the same error when I run the auto_tuning script, could you please tell me how to solve this, sorry to bother you, my friend :)

from badslam.

puzzlepaint avatar puzzlepaint commented on June 28, 2024

As also commented in the other thread, please try setting the BUILD_WITH_MARCH_NATIVE CMake option to ON in your build of g2o.

from badslam.

HuangTY96 avatar HuangTY96 commented on June 28, 2024

Hi, my friend,
thanks for your help, I can run badslam now, but not all dataset on your website, for example, if I run dataset desk_2 and mannequin_4, it will give me totally wrong trajetory, and a lot of warning, If I select "enable loop detection", it will give me an error.(it is still this error:
int g2o::csparse_extension::cs_cholsolsymb(const cs_di, number_t, const cs_dis*, number_t*, int*): cholesky failed!
Cholesky failure, writing debug.txt (Hessian loadable by Octave)**)
I'm confused, is it about my GPU? My environment is Ubuntu16.04 with GTX1060 and CUDA9.1
Selection_002

from badslam.

puzzlepaint avatar puzzlepaint commented on June 28, 2024

See https://www.eth3d.net/slam_benchmark : both for the desk_2 and mannequin_4 datasets, none of the methods seemed to give a good result, so failure on these datasets is expected.

from badslam.

HuangTY96 avatar HuangTY96 commented on June 28, 2024

See https://www.eth3d.net/slam_benchmark : both for the desk_2 and mannequin_4 datasets, none of the methods seemed to give a good result, so failure on these datasets is expected.

Ok, thanks. So could you tell me the reason they fail. Now I want to record some data using my own RGB-D camera, and run these data on badslam, I can make the form of my data just like what we use on badslam, but I don't know how to make sure it won't fail, so could you give me some advice about this?

from badslam.

puzzlepaint avatar puzzlepaint commented on June 28, 2024

Some of the most common failure reasons are:

  • Camera views that don't fully constrain the camera pose (or with too little information for badslam to use it successfully), for example planar walls without much texture
  • Too sparse depth images given by the depth camera
  • Too fast camera movement (can be improved to some extent by increasing the keyframe frequency)
  • Moving objects in the camera view (badslam assumes a static scene)

from badslam.

jsYangCode avatar jsYangCode commented on June 28, 2024

@HuangTY96 hi, I have installed all the dependency, cmake -DCMAKE_BUILD_TYPE=RelWithDebInfo -DCMAKE_CUDA_FLAGS="-arch=sm_61" ..
it shows "CMake did not find CUDA, which is required to build this program.
" in terminal. have u encounter this error?? thank you

from badslam.

Related Issues (20)

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.