Why RANSAC? Because we want a model of good feature matches. - Inliers: - Good matches - Outlier: - Bad matches Interest points (500/image) (640x480) What other algorithms are available? - Exhaustive search - for each featfure, compare all other feactures in other images - Hashing - compute a short descriptor from each feature vector - Nearest neighbor techniques: - k-trees and their variants Putative correspondences (268) (Best match,SSD<20 span=""> Outliers (117) (t=1.25 pixel; 43 iterations) How about outlier rejection? - use SSD (patch1, patch2) > threshold - 1-NN: SSD of the closest matches How to handle too many outliers? RANSAC loop: Select four feature pairs (at random) Compute homography H (exact) Compute inliers where SSD(pi’, H pi) < ε Keep largest set of inliers Re-compute least-squares H estimate on all of the inliers Final inliers (262)
About State of the Art in Artificial Intelligence and Machine Vision