Author: Anders

Due to its random nature, standard RANSAC is not always able to find the optimal set (all inliers)  even for moderately contaminated sets and it is known to perform badly when the number of inliers is less than 50%. But this algorithm is capable of finding the optimal set (hence it is "almost" deterministic) even … Continue reading OptimalRANSAC

These two papers propose clustering as a deterministic alternative to RANSAC. The corresponding pairs of points can be rewritten as  points in 2D space. The inliers will be found close to each other while the outliers will be found further away from the cluster, depending on how well the points correspond. The first paper describes … Continue reading Clustering as an Alternative to RANSAC

This paper propose a deterministic alternative to RANSAC that is rotation and scale invariant, but to a smaller extent large perspective differences. We used this in our first word spotter to remove word outliers. The idea is to compare each point in one image with all the ones in the other image. The relative distance … Continue reading Putative Match Analysis (PUMA)

This paper investigates the use of the spinor tensor as the base for a key point detector, very much like the structure tensor and the Hessian is used for the same purpose. It is shown that the determinant of the Spinor tensor can be approximated by the square of the determinant of Hessian. This is … Continue reading The Spinor Tensor

This paper extends the Harris corner detector by investigating the scale space dimension, i.e. the trace of the Hessian matrix. The point found will be close to the one of the Harris corner detector, but still different. There seem to be situations when this detector is more robust than the Harris, but the computational cost … Continue reading The Extended Structure Tensor with a Scale Space Parameter