The openmvs-quadtree-priors from 3dom-fbk

No issue found under https://github.com/3DOM-FBK/openMVS，CGAL library version

Hello, because I didn't find the issue under https://github.com/3DOM-FBK/openMVS, I am asking here. I was wondering what version of the CGAL library you are using? I reported a lot of CGAL library syntax errors, suspected to be the version reason. Looking forward to your reply.

Why the prior normal vector is not considered in the cost function formula of the paper?

Hello, I read your latest paper: "Multiple View Stereo with quadtree-guided priors", from which I benefited a lot.

However, I am a little confused about why the prior normal vector is not considered in the cost function formula of the paper.

This is also the case in your last paper (Semantically Derived Geometric Constraints for MVS
Reconstruction of Textureless Areas).

Is it because the prior normal vector is not accurate enough? Or some other reason?

I am very interested in your work, and thank you very much for sharing your code, it helps me a lot to learn!! Looking forward to your reply!

Wrong time to generate the quadtree split graph：The inl. HPP: 1481: error: (- 215: an Assertion failed) DataType< _Tp> ::channels == m.channels() || m.empty() in function 'operator='

Hello! Thank you very much for open source this code for us to learn.
But I ran into the following problem when generating the quadtree split graph:

Generating Priors...
*** quadTree.max_level: 10
terminate called after throwing an instance of 'cv::Exception'
What () : OpenCV (3.4.13)/usr/local/include/opencv2 / core/mat. The inl. HPP: 1481: error: (- 215: an Assertion failed) DataType&lt; _Tp&gt; ::channels == m.channels() || m.empty() in function 'operator='

Is there an opencv version conflict? What is your opencv version?
Based on the error, I locate the location of my opencv error file:

template<typename _Tp> inline
Mat_<_Tp>& Mat_<_Tp>::operator = (const Mat& m)
{
    if (m.empty())
    {
        release();
        return *this;
    }
    if( traits::Type<_Tp>::value == m.type() )
    {
        Mat::operator = (m);
        return *this;
    }
    if( traits::Depth<_Tp>::value == m.depth() )
    {
        return (*this = m.reshape(DataType<_Tp>::channels, m.dims, 0));
    }
    CV_Assert(DataType<_Tp>::channels == m.channels() || m.empty());
    m.convertTo(*this, type());
    return *this;
}

Running log:

xzx@xzx-legion-ren7000p:~/xu/openMVS-quadtree-priors/run$ python3 run.py /home/xzx/data/pipes /home/xzx/xu/openMVS-quadtree-priors/run/pipes
Using input dir  :  /home/xzx/data/pipes
     output_dir  :  /home/xzx/xu/openMVS-quadtree-priors/run/pipes
1. Densify point cloud
15:46:02 [App     ] Build date: Jan 17 2023, 13:03:36
15:46:02 [App     ] CPU: AMD Ryzen 7 5800 8-Core Processor               (16 cores)
15:46:02 [App     ] RAM: 23.41GB Physical Memory 65.37GB Virtual Memory
15:46:02 [App     ] OS: Linux 5.4.0-136-generic (x86_64)
15:46:02 [App     ] SSE & AVX compatible CPU & OS detected
15:46:02 [App     ] Command line: --input-file /home/xzx/data/pipes/scene.mvs -w /home/xzx/xu/openMVS-quadtree-priors/run/pipes/mvs -o /home/xzx/xu/openMVS-quadtree-priors/run/pipes/mvs/scene_dense.mvs --verbosity 2 --fusion-mode 0 --max-resolution 6400 --min-resolution 100 --estimate-normals 2 --number-views 10 --filter-point-cloud 0 --resolution-level 1 --number-views-fuse 2 --use-semantic 1 --max-threads 32
15:46:02 [App     ] Camera model loaded: platform 0; camera  0; f 0.551x0.551; poses 14
15:46:02 [App     ] Image loaded   0: DSC_0647.JPG
15:46:02 [App     ] Image loaded   1: DSC_0646.JPG
15:46:02 [App     ] Image loaded   2: DSC_0645.JPG
15:46:02 [App     ] Image loaded   3: DSC_0644.JPG
15:46:03 [App     ] Image loaded   4: DSC_0643.JPG
15:46:03 [App     ] Image loaded   5: DSC_0642.JPG
15:46:03 [App     ] Image loaded   6: DSC_0641.JPG
15:46:03 [App     ] Image loaded   7: DSC_0640.JPG
15:46:03 [App     ] Image loaded   8: DSC_0639.JPG
15:46:03 [App     ] Image loaded   9: DSC_0638.JPG
15:46:03 [App     ] Image loaded  10: DSC_0637.JPG
15:46:03 [App     ] Image loaded  11: DSC_0636.JPG
15:46:03 [App     ] Image loaded  12: DSC_0635.JPG
15:46:03 [App     ] Image loaded  13: DSC_0634.JPG
15:46:03 [App     ] Scene loaded from interface format (568ms):
	14 images (14 calibrated) with a total of 343.89 MPixels (24.56 MPixels/image)
	2473 points, 0 vertices, 0 faces
15:46:04 [App     ] Preparing images for dense reconstruction completed: 14 images (655ms)
15:46:04 [App     ] Selecting images for dense reconstruction completed: 14 images (0ms)

Generating Priors... 
*** quadTree.max_level: 10
terminate called after throwing an instance of 'cv::Exception'
  what():  OpenCV(3.4.13) /usr/local/include/opencv2/core/mat.inl.hpp:1481: error: (-215:Assertion failed) DataType<_Tp>::channels == m.channels() || m.empty() in function 'operator='

Looking forward to your reply!! Thanks!

how to fill each quadtree block with a random color?

Hello, I would like to ask you how to fill each quadtree block with a random color? I'm new to this field, and I don't quite understand quadtrees, can you give me some pointers? Your reply will affect my life, Please!

3dom-fbk / openmvs-quadtree-priors Goto Github PK

openmvs-quadtree-priors's People

Contributors

Stargazers

Watchers

Forkers

openmvs-quadtree-priors's Issues

No issue found under https://github.com/3DOM-FBK/openMVS，CGAL library version

Why the prior normal vector is not considered in the cost function formula of the paper?

Wrong time to generate the quadtree split graph：The inl. HPP: 1481: error: (- 215: an Assertion failed) DataType< _Tp> ::channels == m.channels() || m.empty() in function 'operator='

how to fill each quadtree block with a random color?

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent

Jobs