Centroid

About: Centroid is a research topic. Over the lifetime, 4110 publications have been published within this topic receiving 53637 citations. The topic is also known as: barycenter (geometry) & geometric center of a plane figure.

A normalized quadtree representation

Abstract: The quadtree is a hierarchical data structure used for the compact representation of two-dimensional images. It has the advantages of data compression and fast execution of certain feature computations. However, the quadtree representation of an object is heavily affected by its location, orientation, and relative size. If the image of an object is normalized with respect to its centroid and principal axes, and scaled to a standard size, then its quadtree representation is invariant to rotation, translation, and size change. In addition, the principal moments of objects can be computed from this normalized quadtree representation. These features make it an information-preserving shape descriptor and suitable for tasks such as object recognition.

Method and system for clustering identified forms

Abstract: A method is provided for organizing a plurality of documents that include forms. An initial set of clusters is defined for the plurality of documents. The initial set of clusters is reclustered based on similarity values calculated in multiple feature spaces. For example, a first feature space may be associated with a content of a document while a second feature space may be associated with a content of a form associated with the document. Each cluster has an associated centroid vector in each feature space that is used to represent the cluster. The similarity between the document and each cluster is calculated in both feature spaces. Each document is assigned to the cluster whose centroid is most similar. The cluster centroids may be recalculated and the process repeated until the cluster assignments become stable.

Uncertain Centroid based Partitional Clustering of Uncertain Data

Abstract: Clustering uncertain data has emerged as a challenging task in uncertain data management and mining. Thanks to a computational complexity advantage over other clustering paradigms, partitional clustering has been particularly studied and a number of algorithms have been developed. While existing proposals differ mainly in the notions of cluster centroid and clustering objective function, little attention has been given to an analysis of their characteristics and limits. In this work, we theoretically investigate major existing methods of partitional clustering, and alternatively propose a well-founded approach to clustering uncertain data based on a novel notion of cluster centroid. A cluster centroid is seen as an uncertain object defined in terms of a random variable whose realizations are derived based on all deterministic representations of the objects to be clustered. As demonstrated theoretically and experimentally, this allows for better representing a cluster of uncertain objects, thus supporting a consistently improved clustering performance while maintaining comparable efficiency with existing partitional clustering algorithms.

Shape from texture: the homogeneity hypothesis

Abstract: The authors show how a maximum likelihood estimator can be constructed for homogeneous texture. The estimator turns out to be strikingly simple. It simply enforces, iteratively, coincidence of the centroid of the observed elements with the centroid of the imaging window. Whereas the estimator is based on first moments, error bounds are obtained, the authors show, from second moments. Experiments indicate that, for real textures, orientation estimates fall well inside predicted bounds. >