jimjibone / robotarm Goto Github PK

What is it?!?! Just like the RANSAC one. Mine is present wrapping or whatever it's called, Jarvis' March possibly. What do they use?

What is it?!?!?!? My current one is below, is their one the same?

1. Randomly select 'n' points to be possible inliers.
2. Find all the points (other points) that are not possible inliers.
3. For each set of 3 possible inliers:
4. Find the plane equation for these 3 points.
5. If the points are valid add them to the planes array.
6. Iterate through all the valid planes and determine their confidence by comparing them to the other points.
7. Get the current plane.
8. For each other point find the distance from point to plane.
9. If the point is within the tolerance from the plane then add 1 to the plane's confidence. If it is outside the confidence then remove 1 from the plane's confidence.
10. Find the plane with the highest confidence value.
11. Add this plane to the confident planes array.
12. Sort the confident planes array by highest to lowest confidence.
13. The most confident plane fo the current frame (iteration) can be found at index 0 of the confident planes array.