Video target tracking method based on manifold particle filter algorithm

Abstract

The invention discloses a video target tracking method based on a manifold particle filter algorithm. According to the method, projective transformation of a video image is constructed to a matrix lie group, the projective transformation parameter of a target is used as state variable, and a state transition model is established on the lie group; a target area in the video image is described by means of covariance description; a state sample is extracted along a manifold geodesic by means of the particle filter algorithm on the lie group; an internal implication mean value is solved by means of the internal implication Gaussian Newton arithmetic, state estimation of a system is obtained, and target tracking is achieved. According to the video target tracking method based on the manifold particle filter algorithm, effect of the noise statistical property of the Euclidean space on weight variance is reduced, solving of the particle degenerating problem is benefited, and tracking accuracy and robustness of the algorithm are improved.

Description

technical field [0001] The invention relates to a visual target tracking system, which uses a particle filter method on a manifold to track a target in a video image, and belongs to the technical field of nonlinear system filtering and visual image processing. Background technique [0002] Visual tracking is the key technology to realize intelligent monitoring. It integrates advanced technologies in several fields such as image processing, pattern recognition, artificial intelligence, automatic control and computer, and is used in military visual guidance, video surveillance, robot visual navigation, medical diagnosis and meteorological analysis. etc. are widely used. [0003] Currently commonly used visual tracking methods can be generally divided into five types: area-based tracking, dynamic contour-based tracking, feature-based tracking, model-based tracking, and motion estimation-based tracking. The most commonly used method is to establish geometric parameter models su...

AI Technical Summary

Problems solved by technology

[0005] Although the particle filter can be applied to all nonlinear and non-Gaussian systems, it is not limited by the nature of the noise, but the existing particle filter algorithms are all carried out in the Euclidean space. When the particle filter algorithm is used to track high-dimensional systems and multi-targets , the same problem of "curse of dimensionality"

[0006] When performing visual tracking, the covariance of the observation noise is likely to be unknown and time-varying, or when the covariance matrix is ​​used to express the target region in the image, when performing image matching, it is necessary to calculate the difference between the covariance matrices of the two image regions , since the covariance is a positive definite matrix, all positive definite matrices form a Riemannian manifold, so it is not suitable to use the Euclidean space method to track at this time, and the space differential geometric characteristics of the positive definite matrix must be used to construct a more effective algorithm

Images