Wyniki wyszukiwania - BazTech

1

A data association model for analysis of crowd structure

Zitouni M. Sami, Śluzek Andrzej

International Journal of Applied Mathematics and Computer Science

|

2022

|

Vol. 32, no. 1

81--94

EN

The paper discusses a non-deterministic model for data association tasks in visual surveillance of crowds. Using detection and tracking of crowd components (i.e., individuals and groups) as baseline tools, we propose a simple algebraic framework for maintaining data association (continuity of labels assigned to crowd components) between subsequent video-frames in spite of possible disruptions and inaccuracies in tracking/detection algorithms. Formally, two alternative schemes (which, in practice, can be jointly used) are introduced, depending on whether individuals or groups can be prospectively better tracked in the current scenario. In the first scheme, only individuals are tracked, and the continuity of group labels is inferred without explicitly tracking the groups. In the second scheme, only group tracking is performed, and associations between individuals are inferred from group tracking. The associations are built upon non-deterministic estimates of memberships (individuals in groups) and estimates obtained directly from the baseline detection and tracking algorithms. The framework can incorporate any detectors and trackers (both classical or DL-based) as long as they can provide some geometric outlines (e.g., bounding boxes) of the crowd components. The formal analysis is supported by experiments in sample scenarios, where the framework provides meaningful performance improvements in various crowd analysis tasks.

2

Non-linear multimodal object tracking based on 2d lidar data

Thuy M., Puente León F.

Metrology and Measurement Systems

|

2009

|

Vol. 16, nr 3

359-369

EN

The contribution introduces a novel approach for tracking objects based on two-dimensional lidar data. As a central tracking engine, we employ a particle-filter-based solution which is capable of modelling non-linear dynamic processes as well as non-Gaussian noise distributions for the underlying process and sensor as well. In contrast to other lidar-based tracking approaches, no newly detected objects have to be associated to already known objects in an explicit manner. Since our weighting function is multi-modal, the association is done by the filter itself.

3

GA-JPDA and fuzzy data association algorithms coupled with IMM-PF estimator for highly maneuvering multiple-target tracking

Djouadi M. S., Morsly Y.

Archives of Control Sciences

|

2007

|

Vol. 17, no. 3

311-327

EN

In this paper, we present filtering algorithm to perform accurate estimation in jump Markov nonlinear systems, in case of multi-target tracking. With this paper, we aim to contribute in solving the problem of model-based body motion estimation by using data coming from visual sensors. The Interacting Multiple Model (IMM) algorithm is specially designed to track accurately targets whose state and/or measurement (assumed to be linear) models change during motion transition. However, when these models are nonlinear, the IMM algorithm must be modified in order to guarantee an accurate tracking. In order to deal with this problem, the IMM algorithm was combined with the Unscented Kalman Filter (UKF). Even if the later algorithm proved its efficacy in nonlinear model case, it presents a serious drawback in the case of non Gaussian noise. To deal with this problem we propose to substitute the UKF with the Particle Filter (PF). To overcome the problem of data association, we propose the use the JPDA approach. To reduce the computational burden of this technique, we choose firstly the most likely feasible events by applying a Genetic Algorithm; finally the derived algorithm from the combination of the IMM-PF algorithm and the GA-JPDA approach is noted GA-JPDA-IMM-PF. To insure a more reduction of the computation complexity of the latter data association approach, we propose a fuzzy data association approach which we combine with the IMM-PF estimator, the derived algorithm is noted fuzzy IMM-PF. Finally the two algorithms are compared according to the target loss rate inferred by each of them.