Motion characterization from co-occurrence vector descriptor

Authors:
Zeev Zalevsky;Ehud Rivlin;Michael Rudzsky
Affiliations:
Department of Computer Science, Technion Israel Institute of Technology, Technion City, 32000 Haifa, Israel;Department of Computer Science, Technion Israel Institute of Technology, Technion City, 32000 Haifa, Israel;Department of Computer Science, Technion Israel Institute of Technology, Technion City, 32000 Haifa, Israel
Venue:
Pattern Recognition Letters
Year:
2005

Citing 4
Cited 0

Detecting texture periodicity from the co-occurrence matrix

Pattern Recognition Letters
Estimating the heading direction using normal flow

International Journal of Computer Vision
Multidimensional co-occurrence matrices for object recognition and matching

Graphical Models and Image Processing
Motion Recognition Using Nonparametric Image Motion Models Estimated from Temporal and Multiscale Cooccurrence Statistics

IEEE Transactions on Pattern Analysis and Machine Intelligence

Quantified Score

Hi-index	0.10

Visualization

Abstract

A novel operator called the co-occurrence vector to be used for motion characterization is presented. The vector has the advantage of requiring low computation complexity. It allows estimating the relative position, the scale and the rotation of objects in the scene image. This estimation may be useful for obtaining time to impact and relative distance between objects for navigation and collision avoidance. Experiments confirm the value of co-occurrence vector for estimation of motion parameters and as a tool for characterization of scale, rotation and relative shift of objects in the scene.