Abstract:Visual tracking can be formulated as a state estimation problem of target representation based on observations in image sequences. To investigate the integration of rough models from multiple cue and to explore computationally efficient algorithms, this paper formulates the problem of multiple cue integration and tracking to combine Interactive Multiple Model (IMM) with particle filter (IMM_PF). Interactive Multiple Model can estimate the multiple cue state of a dynamic system with several behavioral models that switch from one to another using model likelihoods and model transition probabilities. For the problem of visual tracking, the model of IMM is adopted to three target observation models: Corrected Background Weighted Histogram (CBWH), Completed Local Ternary Patterns (CLTP) and Histogram of Oriented Gradients (HOG). The probabilities of these models are corresponding to the weights of multiple cues. IMM_PF then dynamically adjusts the weights of different features. Compared with those state-of-the-art methods in the tracking literature, this algorithm can track the object accurately in conditions of rotation, abrupt shifts, as well as clutter and partial occlusions occurring to the tracking object with good robustness, as demonstrated by experimental results.

Visual tracking with adaptive multi-cue fusion particle filter

Visual Object Tracking Based on Motion-Adaptive Particle Filter under Complex Dynamics

Particle Filter Based Visual Tracking with Multi-Cue Adaptive Fusion

The improved particle filter for object tracking

Visual tracking using particle filters based on data fusion

Particle filter with affine transformation for multiple key points tracking

Efficient Particle Filter and Its Application in Contour Tracking

Dynamic Multi-Cue Tracking Using Particle Filter

Probabilistic Multiple Cue Integration for Particle Filter Based Tracking

An Adaptive Fusion Strategy Based Multiple-Cue Tracking

Visual Tracking Via Adaptive Interactive Fusion

Robust Visual Tracking Base on Adaptively Multi-Feature Fusion and Particle Filter

Object Tracking Based on Multi-Information and Auto-Adapted Particle Filter

Object contour tracking using multi-feature fusion based particle filter

Adaptive Tracking Algorithm Based on Particle Grouped Optimization

Multi-target Tracking Based on Fusion-Feedback Particle Filter

Particle Filtering Algorithm Based on Dynamic Multi-Feature Fusion

Dim and Small Target Tracking Using an Improved Particle Filter Based on Adaptive Feature Fusion

Robust visual tracking based on interactive multiple model particle filter by integrating multiple cues

Robust Head Tracking with Particles Based on Multiple Cues Fusion

Robust Visual Tracking Based on Joint Multi-Feature Histogram by Integrating Particle Filter and Mean Shift