This paper proposes a new 3D modeling technique based on feature points using spatio-temporal relationship. Normally, the generation of a 3D model from a real scene requires the computation of the depth of the model vertices from a dense correspondence map of the whole image. This is very time consuming, plus it is also quite difficult to achieve an accurate depth. The proposed method can gener...

This paper presents computationally efficient object detection, matching and categorization via Agglomerative Correspondence Clustering (ACC). We implement ACC for feature correspondence and object-based image matching exploiting both photometric similarity and geometric consistency from local invariant features. Objectbased image matching is formulated here as an unsupervised multi-class clust...

Abstact Point/feature matching in 2D images has been well studied in the computer vision community. With the huge advancement in shape scanning and acquisition technology, feature matching has recently found usefulness in 3D shape matching and searching. We study the general problem of feature correspondence in 3D shapes represented by triangular meshes. This project is a survey and comparison ...

We present a statistical decision approach for the point set matching of unordered feature sets. Both feature sets have associated uncertainties, and the number of elements in each set may be different. Computation of the match likelihood requires a correspondence between feature sets; we solve the correspondence problem in polynomial time using the Hungarian algorithm. We also consider the pro...

The ability to extract and follow time-varying features in volume data obtained from large-scale numerical simulations makes it possible for scientists to see and validate the modeled phenomena and processes because the extracted features generally take much less space to store and less computing resources to visualize. Previous feature extraction and tracking methods extract features in each t...

The research reported here addresses the problem of detecting and tracking independently moving objects from a moving observer in real time, using corners as object tokens. Local image-plane constraints are employed to solve the correspondence problem removing the need for a 3D motion model. The approach relaxes the restriciive static-world assumption conventionally made, and is therefore capab...