Back to results
US 2007/0183629 A1Published application
Method for tracking objects in videos using covariance matrices
AssigneeMitsubishi Electric Research Laboratories, Inc.
InventorsFatih Porikli, Oncel Tuzel
Life of this patent
Estimating the term…
Abstract
A method is provided for tracking a non-rigid object in a sequence of frames of a video. Features of an object are extracted from the video. The features include locations of pixels and properties of the pixels. The features are used to construct a covariance matrix. The covariance matrix is used as a descriptor of the object for tracking purposes. Object deformations and appearance changes are managed with an update mechanism that is based on Lie algebra averaging.
Claims26
A computerized method for tracking an object in a sequence of frames, comprising the steps of:
- extracting a feature frame from each frame in a sequence of frames acquired of an object;
- constructing a set of feature vectors for each feature frame, each feature vector corresponding to a different window of features in the feature frame;
- determining a corresponding set of covariance matrices from the set feature vectors;
- measuring distances between each covariance matrix in the set of covariance matrices and a model covariance matrix representing the object;
- and
- selecting a particular window as a location of the object in the frame, the particular window having a corresponding covariance matrix with a minimum distance from the model covariance matrix.
3 more elements
Dependencies are inferred from the claim text.