Stack Overflow
Stack Overflow

Questions tagged [structure-from-motion]

72 questions
20
votes
1 answer

In computer vision, what does MVS do that SFM can't?

I'm a dev with about a decade of enterprise software engineering under his belt, and my hobbyist interests have steered me into the vast and scary realm of computer vision (CV). One thing that is not immediately clear to me is the division of labor…
smeeb
  • 27,777
  • 57
  • 250
  • 447
11
votes
1 answer

What's the conceptual difference between bundle adjustment and structure from motion?

In my mind they both mean reconstructing 3D coordinates from matched points in 2D images. What's the difference between these concepts and multi-view stereo? Which one do you call an algorithm that computes a sparse point cloud from keypoint…
icguy
  • 560
  • 4
  • 14
7
votes
3 answers

Is the recoverPose() function in OpenCV is left-handed?

I run simple test for OpenCV camera pose estimation. Having a photo and the same photo scaled up (zoomed in) I use them to detect features, calculate essential matrix and recover camera poses. Mat inliers; Mat E = findEssentialMat(queryPoints,…
vagran
  • 867
  • 1
  • 7
  • 18
7
votes
4 answers

Camera pose estimation: How do I interpret rotation and translation matrices?

Assume I have good correspondences between two images and attempt to recover the camera motion between them. I can use OpenCV 3's new facilities for this, like this: Mat E = findEssentialMat(imgpts1, imgpts2, focal, principalPoint, RANSAC, 0.999,…
oarfish
  • 4,116
  • 4
  • 37
  • 66
5
votes
1 answer

How do I estimate positions of two cameras in OpenCV?

I have two sets of corresponding points from two images. I have estimated the Essential matrix which encodes the transformation between the cameras: E, mask = cv2.findEssentialMat(points1, points2, 1.0) I've then extracted the rotation and…
nickponline
  • 25,354
  • 32
  • 99
  • 167
4
votes
1 answer

How do ARCore or ARKit produce real-time augmentations of live video?

So a while back about a year ago I was interested in building my own barebones augmented reality (AR) library. My goal was to be able to take a video of something (anything really) and then be able to place augmentations (3D objects that weren't…
smeeb
  • 27,777
  • 57
  • 250
  • 447
4
votes
0 answers

How do you estimate 3D points from two images point correspondence?

I'm trying to implement a basic structure-from-motion pipeline in Python using OpenCV to generate a point cloud from two images given the corresponding point in both images. I manage to get 3D points out of it but the locations don't really make…
nickponline
  • 25,354
  • 32
  • 99
  • 167
4
votes
1 answer

Structure from Motion with Optical Flow

Let say I have a video from a drive recorder. I want to construct the recorded scene's points cloud using structure from motion technique. First I need to track some points. Which algorithm can yield a better result? By using the sparse optical flow…
Hafiz Hilman Mohammad Sofian
  • 1,503
  • 5
  • 21
  • 37
3
votes
1 answer

Why do triangulated points not project back to same image points in OpenCV?

I have two corresponding image points (2D) visualized by the same camera with intrinsic matrix K each coming from different camera poses (R1, t1, R2, t2). If I triangulate the corresponding image points to a 3D point and then reproject it back to…
nickponline
  • 25,354
  • 32
  • 99
  • 167
3
votes
1 answer

what track and track length means in SFM

when I read the paper "Structure-from-Motion Revisited". I met a few words I didn't understand，　 "track" and "track length". what exactly these terms mean?
justPassBy
  • 87
  • 5
3
votes
2 answers

Structure From Motion From Multiple Views w/ MATLAB

I am a novice when it comes to the topic of Structure From Motion. I have been trying to follow the tutorial here in the MathWorks webpage for SFM: LINK. However, after running the code, I get this error message: Warning: Maximum number of trials…
troymyname00
  • 670
  • 1
  • 14
  • 32
3
votes
0 answers

BoofCV Structure from motion (Sfm) by Example

I am playing around with structure from motion (Sfm) and 3D model reconstruction for the first time. The end result is to be able to generate a (fairly accurate) 3D model from a series of images. Being a JVM developer, my natural inclination was to…
smeeb
  • 27,777
  • 57
  • 250
  • 447
3
votes
1 answer

Simple Orthographic Structure from Motion using R -- Determining Metric Constraints

I would like to build a simple structure from motion program according to Tomasi and Kanade [1992]. The article can be found below:…
Alex Witsil
  • 855
  • 8
  • 22
2
votes
1 answer

How to calculate the relative scale from two fundamental matrices that share a camera view?

I understand that the scale represented by the fundamental matrix FAB between two camera views A and B is only correct up to a scale (ie, you don't know if you're for example looking at small trees up close or larger trees farther away). However,…
alvion
  • 1,963
  • 3
  • 15
  • 23
2
votes
0 answers

How to use own features, computed in openCV, in visualSFM pipeline

I try to do a 3D reconstruction from multiple images. I am currently using the visual SFM Pipeline that takes in Lowe SIFT features as .feat and .mat format. Both are binary so I could not read them with editor. Due to the documentation of visual…
Max Krappmann
  • 490
  • 5
  • 19
1
2 3 4 5