Straightforward Adaptation of Particle Filter to Fish Eye Images for Top View Pedestrian Tracking

Hicham Talaoubrid (1, 2) , Khizar Hayat (3) , Baptiste Magnier (2)

1 L2TI - Laboratoire de Traitement et Transport de l'Information
2 Euromov DHM - EuroMov - Digital Health in Motion
3 University of Nizwa

Object tracking in computer vision, essential for detecting, locating, and predicting object movement in videos [5], faces challenges with fisheye cameras due to wide-angle views, distortion, and complex perspective shifts. Particle Filters offer a solution by providing a flexible framework for tracking objects in nonlinear, wide-angle scenes. They excel in handling fisheye image characteristics like distortion correction and probabilistic state estimation, making them valuable for robust object tracking in fisheye camera setups.

Mots clés: Fisheye camera; Particle filter; Histograms; Object tracking

Domaines: Traitement du signal et de l'image [eess.SP]

Hicham Talaoubrid, Khizar Hayat, Baptiste Magnier. Straightforward Adaptation of Particle Filter to Fish Eye Images for Top View Pedestrian Tracking. ICASSP 2024 - 2024 IEEE International Conference on Acoustics, Speech and Signal Processing, Apr 2024, Seoul, South Korea. IEEE, 2024. ⟨hal-04554295⟩