Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Does anyone know why is this happening #505

Closed
IvanMN7 opened this issue Jun 11, 2019 · 5 comments
Closed

Does anyone know why is this happening #505

IvanMN7 opened this issue Jun 11, 2019 · 5 comments

Comments

@IvanMN7
Copy link

IvanMN7 commented Jun 11, 2019

With the SFM activated:

ProblemaMeshroom1

And this with the texture activated:

ProblemaMeshroom2

As seen in the Structure points, the object to which I took photos appears the structure with the points but at the time of doing the "load model" and with the texture does not appear the object, is broken.

Any solution?

@natowi
Copy link
Member

natowi commented Jun 11, 2019

Your sheep has few destinct features. The table/chair consumes 90% of your image, so most of your sheep (the areas with low maches) is filtered out.
Take some closeup images and try again. The object of interest should fill most of the image frame.

@IvanMN7
Copy link
Author

IvanMN7 commented Jun 12, 2019

Your sheep has few destinct features. The table/chair consumes 90% of your image, so most of your sheep (the areas with low maches) is filtered out.
Take some closeup images and try again. The object of interest should fill most of the image frame.

I did what you told me and now I have another problem:

When I try to insert some images without background in SFM(StructureFromMotion) section this error apears me:
Aborted (core dumped)

This is the Log:

Program called with the following parameters:

  • describerTypes = "sift"
  • extraInfoFolder = "/home/nube7/Escritorio/Meshroom/robot3(studio)/MeshroomCache/StructureFromMotion/4f06d43ece206e6f3c753b1fee1fa555fa7c3f7f"
  • featuresFolders = = [/home/nube7/Escritorio/Meshroom/robot3(studio)/MeshroomCache/FeatureExtraction/fa4d2043c60d9ed83166e317d360b5b85f8dba18]
  • initialPairA = ""
  • initialPairB = ""
  • input = "/home/nube7/Escritorio/Meshroom/robot3(studio)/MeshroomCache/CameraInit/2b0316e9aa38fd9274065bcd14f3d389db0b2b9e/cameraInit.sfm"
  • interFileExtension = ".abc"
  • localBAGraphDistance = 1
  • localizerEstimator = Unknown Type "N11aliceVision16robustEstimation16ERobustEstimatorE"
  • localizerEstimatorError = 0
  • localizerEstimatorMaxIterations = 4096
  • lockAllIntrinsics = 0
  • lockScenePreviouslyReconstructed = 0
  • matchesFolders = = [/home/nube7/Escritorio/Meshroom/robot3(studio)/MeshroomCache/FeatureMatching/8f6d0684673917cd64e90a40a32aed72b683b7e1]
  • maxAngleInitialPair = 40
  • maxNumberOfMatches = 0
  • maxReprojectionError = 4
  • minAngleForLandmark = 2
  • minAngleForTriangulation = 3
  • minAngleInitialPair = 5
  • minInputTrackLength = 2
  • minNumberOfObservationsForTriangulation = 2
  • output = "/home/nube7/Escritorio/Meshroom/robot3(studio)/MeshroomCache/StructureFromMotion/4f06d43ece206e6f3c753b1fee1fa555fa7c3f7f/sfm.abc"
  • outputViewsAndPoses = "/home/nube7/Escritorio/Meshroom/robot3(studio)/MeshroomCache/StructureFromMotion/4f06d43ece206e6f3c753b1fee1fa555fa7c3f7f/cameras.sfm"
  • useLocalBA = 1
  • useOnlyMatchesFromInputFolder = 0
  • useRigConstraint = 1
  • useTrackFiltering = 1 (default)
  • verboseLevel = "info"

Loading features
0% 10 20 30 40 50 60 70 80 90 100%
|----|----|----|----|----|----|----|----|----|----|
[11:19:02.885364][info] Supported CUDA-Enabled GPU detected.


[11:19:02.892875][info] Fuse matches into tracks:
- # tracks: 497
- # images in tracks: 10
[11:19:02.892914][info] TrackLength, Occurrence
[11:19:02.892968][info] 2 316
[11:19:02.892987][info] 3 136
[11:19:02.892998][info] 4 21
[11:19:02.893008][info] 5 11
[11:19:02.893019][info] 6 9
[11:19:02.893037][info] 7 4
Automatic selection of an initial pair:
0% 10 20 30 40 50 60 70 80 90 100%
|----|----|----|----|----|----|----|----|----|----|


[11:19:02.918734][info] Initial pair is:
- [A] view id: 1676625766, filepath: /home/nube7/Descargas/Robot 3 (studio)/IMG_20190612_101848.jpg
- [B] view id: 2009016528, filepath: /home/nube7/Descargas/Robot 3 (studio)/IMG_20190612_101845.jpg
[11:19:02.918885][info] 141 matches in the image pair for the initial pose estimation.
[11:19:02.955958][info] Bundle adjustment start.
[11:19:02.956064][info] Start bundle adjustment iteration: 0
block_sparse_matrix.cc:81 Allocating values array with 35424 bytes.
detect_structure.cc:113 Schur complement static structure <2,3,6>.
detect_structure.cc:113 Schur complement static structure <2,3,6>.
[11:19:03.018721][info] Bundle Adjustment Statistics:
- local strategy enabled: no
- adjustment duration: 0.0404906 s
- poses:
- # refined: 2
- # constant: 0
- # ignored: 0
- landmarks:
- # refined: 246
- # constant: 0
- # ignored: 0
- intrinsics:
- # refined: 0
- # constant: 1
- # ignored: 0
- # residual blocks: 492
- # successful iterations: 5
- # unsuccessful iterations: 0
- initial RMSE: 0.73318
- final RMSE: 0.64448
[11:19:03.018879][info] Remove outliers:
- # outliers residual error: 2
- # outliers angular error: 0
[11:19:03.018901][info] Bundle adjustment iteration: 0 took 62 msec.
[11:19:03.018913][info] Start bundle adjustment iteration: 1
[11:19:03.019380][warning] Bundle Adjustment failed, the solution is not usable.
terminate called after throwing an instance of 'std::runtime_error'
what(): Initialization failed after trying all possible initial image pairs.
Aborted (core dumped)

@natowi
Copy link
Member

natowi commented Jun 28, 2019

Looks like you don´t get enough matches with default settings. You could optimize the settings.
You could also add some removable colour to the uniform areas to provide better features for the algorithms.

@hargrovecompany
Copy link

hargrovecompany commented Jul 3, 2019

light reflection on the table top. you can see the different position of flourescent light reflection in every image. software is confused about locations because those features (reflections of light) are not in a constant location. get very close to the sheep so that it takes up most of the image.

@hargrovecompany
Copy link

also, after looking again it appears to me that the sheep might have some reflective surface. So white points of reflected light would be in a different location on every one of the little "curls" of the sheep's wool. And, I'm just guessing, i actually can't tell clearly if that's an issue. If that's the case try setting up several lamps around the table....if it's still getting reflections, rub damp bar soap on your hands and smear all over the object then dust with baking powder.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

3 participants