Lecture-14 - Good Features to Track 1 Good Features to...

Info iconThis preview shows pages 1–11. Sign up to view the full content.

View Full Document Right Arrow Icon
1 Good Features to Track
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
2 Good Features to Track • Corners • Moravec’s Interest Operator Corners = Q y Q y x Q y x Q x f f f f f f C 2 2 = 2 1 0 0 λ C
Background image of page 2
3 Corners For perfectly uniform region If Q contains an ideal step ed if Q contains a corner of black square on white background 0 2 1 = = λ λ 0 , 0 1 2 > = 0 2 1 ! λ≥ Algorithm Corners • Compute the image gradient over entire image f. • For each image point p: – form the matrix C over (2N+1)X(2N+1) neighborhood Q of p; – compute the smallest eigenvalue of C; – if eigenvalue is above some threshold, save the coordinates of p into a list L.
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
4 Algorithm Corners • Sort L in decreasing order of eigenvalues. • Scanning the sorted list top to bottom: for each current point, p, delete all other points on the list which belong to the neighborhood of p. Results
Background image of page 4
5 Results
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
6 Moravec’s Interest Operator Algorithm • Compute four directional variances in horizontal, vertical, diagonal and anti- diagonal directions for each 4 by 4 window. • If the minimum of four directional variances is a local maximum in a 12 by 12 overlapping neighborhood, then that widow (point) is interesting.
Background image of page 6
7
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
8 ∑∑ == + + + + + = 3 0 2 0 2 )) , 1 ( ) , ( ( ji h j y i x P j y i x P V + + + + + = 2 0 3 0 2 )) 1 , ( ) , ( ( v j y i x P j y i x P V + + + + + + = 2 0 2 0 2 )) 1 , 1 ( ) , ( ( d j y i x P j y i x P V + + + + + = 2 0 3 1 2 )) 1 , 1 ( ) , ( ( a j y i x P j y i x P V )) , ( ), , ( ), , ( ), , ( min( ) , ( y x V y x V y x V y x V y x V a d v h = = therwise local y x ifV y x I 0 0 max ) , ( 1 ) , (
Background image of page 8
9 Feature-based Matching
Background image of page 9

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
10 Feature-based Matching • The input is formed by f1 and f2, two frames of an image sequence, and a set of corresponding feature points in two frames. • Let Q1, Q2 and Q’ be three NXN image regions. • Let “d” be the unknown displacement vector between f1 and f2 of a feature point “p”, on which Q1 is centered. Algorithm • (1) Set d=0, center Q1 on p1. • (2) Estimate the displacement “d0” of “p”, center of “Q1”, using Lucas and Kanade method. Let d=d+d0 . • (3) Let Q’ bet the patch obtained by warping Q1 according to “d0”. – Compute Sum of Square (SSD) difference between new patch Q’ and corresponding patch Q2 in frame f2.
Background image of page 10
Image of page 11
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 28

Lecture-14 - Good Features to Track 1 Good Features to...

This preview shows document pages 1 - 11. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online