Nonuniform Image Motion Estimation Using the Maximum a Posteriori Principle

Document Type

Article

Publication Date

1-1-1992

Abstract

This paper is concerned with the establishment of an literative scheme for frame-to-frame motion estimation from a pair of noisy images. The algorithm is developed by assuming that the Karhunen-Loeve (KL) coefficients of the motion vector waveform are zero mean and Gaussian random variables. Following the derivation of the generalized maximum likelihood (GML) algorithm [1], [2] and invoking the maximum a posteriori (MAP) criterion, we arrive at an iterative motion estimator. A linear analysis of the algorithm is presented and the convergence of the algorithm is discussed. Simulation experiments are performed and comparisons are made with the GML algorithm, the algorithm reported by Netravali and Robbins [3], and the scheme developed by Horn and Schunck [4]. © 1992 IEEE

Publication Title

IEEE Transactions on Image Processing

Share

COinS