Advanced wang–landau monte carlo‐based tracker for abrupt motions

Abstract

Conventional tracking methods rarely consider abrupt motions and easily fail to track the abrupt motion of an object because they are based on the assumption of smooth motion. To assuage this problem, we propose a novel tracking algorithm combining the background subtraction method with the Wang-Landau Monte Carlo (WLMC) sampling method for dealing with abrupt motions effectively. First, the visual background extractor (ViBe), a background subtraction technique, is introduced to detect the object roughly. Second, we integrate the density-of-states term and ViBe information into a WLMC-based tracking framework. By employing the density-of states term and ViBe information, the Wang-Landau sampling scheme efficiently tracks the target in the whole state space in a short time. Third, in order to obtain good performance in terms of scalability, an independent fast scale estimation method is applied. Experimental results in both quantitative and qualitative measures demonstrate that our algorithm efficiently samples and tracks the target not only in position but also in scale variance accurately and robustly.

Visual Tracking Using Wang–Landau Reinforcement Sampler

“…However, these methods do not contain a feedback process and cannot reflect the current visual tracking environment and results. Liu et al [16] combined WLMC sampling with a visual background extractor, considerably reducing the state space of the target. They independently dealt with scale changes in the target using a fast scale estimation algorithm.…”

“…where p i (Y t |X t ) is the likelihood at the i-th state, which is defined in (3). In (16) • Reinforcement learning…”