Showing papers on "Scale-invariant feature transform published in 1991"

Reconfigurable Mesh Algorithms for the Hough Transform.

Abstract: We develop parallel algorithms to compute the Hough transform on a reconfigurable mesh with buses (RMESH) multiprocessor. The p angle Hough transform of an N × N image can be computed in O(p log(N/p)) time by an N × N RMESH, in O((p/N) log N) time by an N × N2 RMESH with N copies of the image pretiled, in O((p/[formula]) log N) time by an N1.5 × N1.5 RMESH, and in O((p/N) log N) time by an N2 × N2 RMESH.

Three-dimensional motion analysis and structure recovering by multistage Hough transform

Abstract: A new approach for the detection of motion of three-dimensional rigid bodies from two- dimensional images is presented. The approach is based on two main stages. In the first stage, the positions and velocities of feature points are detected from two-dimensional images. In the second stage, the rotation and the translation velocity of each body is detected from the positions and velocities of the set of feature points. We employ the Hough transform method in both stages. We describe the details of the second stage and the method of computation reduction in Hough transform. The effectiveness of our method is confirmed through computer experiment.© (1991) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Simple object recognition by neural networks: application of the Hough transform

Abstract: A ‘neural-like’ parallel system is described for the real-time implementation of the Hough transform. It is applied to simple object recognition in computer vision.

Line Identification by Hough Transform with Image Subdivision

Abstract: To cope with memory and time requirements of the Hough transform when identifying lines here, a line parametrisation is investigated which allows a relatively small and compact accumulator, a fast algorithm to fill the accumulator and its implementation in parallel hardware is developed and finally,an algorithm to combine the lines identified in subimages is described: by this image subdivision Hough transform and cluster detection can be applied to proportionally smaller accumulators in parallel.

Robust key point matching for dynamic scenes

Abstract: Reliable and well distributed correspondences between sparsely sampled photographic images of dynamic scenes are needed in many computer vision applications including image-based rendering and 3D reconstruction. We propose a method based on local affine models that deals with local motion and resolves ambiguities. For most images this results in a considerably larger set of matches and lower error count than classical matching approaches. Fairly reliable matches can be obtained by applying a key point detector and descriptor like SIFT and assigning to each key point the nearest neighbor in a multidimensional descriptor space. To limit false positives due to ambiguities and non-existent correspondences, the distance ratio between the two nearest neighbors is thresholded and the match has to be found in both directions. While discarding many potentially correct matches, a considerable amount of wrong matches remains (figure 1b). This leads to inconsistencies in multi-image key point tracks, suggesting that multiple points in one image represent the same object point. Since global measures to eliminate the wrong matches on a track cannot easily be derived from the descriptor distances, common solution are to drop all of the inconsistent tracks, to only work on image pairs, or to use underlying global models like a fundamental matrix (figure 1c), or a 3D reconstruction to reject outliers. However, those models cannot handle local motion, considerably reduce the matching density, and cannot always be used to resolve ambiguities (figure 1d).