How does the "Dense Reconstruction" in the "Structure From Motion From Multiple Views" MATLAB example work?

6 views (last 30 days)
I am trying to adapt the example in the MATLAB documentation on Structure From Motion From Multiple Views (https://uk.mathworks.com/help/vision/examples/structure-from-motion-from-multiple-views.html). However, I'm not sure I understand the section on dense reconstruction. Mainly the part shown below where it seems that points from the first view are tracked across all other views:
% Track the points across all views.
for i = 2:numel(images)
% Read and undistort the current image.
I = undistortImage(images{i}, cameraParams);
% Track the points.
[currPoints, validIdx] = step(tracker, I);
% Clear the old matches between the points.
if i < numel(images)
vSet = updateConnection(vSet, i, i+1, 'Matches', zeros(0, 2));
end
vSet = updateView(vSet, i, 'Points', currPoints);
% Store the point matches in the view set.
matches = repmat((1:size(prevPoints, 1))', [1, 2]);
matches = matches(validIdx, :);
vSet = updateConnection(vSet, i-1, i, 'Matches', matches);
end
To create a 3D reconstruction shouldn't you track points from all views, as points in the first image may not be visable in every other image?
Thank you in advance!

Accepted Answer

Sai Bhargav Avula
Sai Bhargav Avula on 31 Jul 2019
For SfM the approach used assumes an ordered sequence of views. The tracks are computed from pairwise point correspondence. The function viewSet maintains the pairwise correspondence and tracks. In the given example updateConnection and updateViews functions clears the old matches and populates with the new matches in the viewSet.
For a detailed understanding of the viewSet function refer the following link:

More Answers (0)

Categories

Find more on Filter Banks in Help Center and File Exchange

Products


Release

R2019a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!