A Bank of Unscented Kalman Filters for Multimodal Human Perception with Mobile Service Robots

Abstract

A new generation of mobile service robots could be ready soon to operate in human environments if they can robustly estimate position and identity of surrounding people. Researchers in this field face a number of challenging problems, among which sensor uncertainties and real-time constraints. In this paper, we propose a novel and efficient solution for simultaneous tracking and recognition of people within the observation range of a mobile robot. Multisensor techniques for legs and face detection are fused in a robust probabilistic framework to height, clothes and face recognition algorithms. The system is based on an efficient bank of Unscented Kalman Filters that keeps a multi-hypothesis estimate of the person being tracked, including the case where the latter is unknown to the robot. Several experiments with real mobile robots are presented to validate the proposed approach. They show that our solutions can improve the robot's perception and recognition of humans, providing a useful contribution for the future application of service robotics.

A framework for the design of person following behaviors for social mobile robots

“…Due to their complementarity, most of the recent studies [8], [9], [10], [11] have focused on the fusion of the information provided by these two sensors: laser sensors provide accurate distance estimation while the cameras provide discriminative information.…”

A Multi-Modal Unscented Kalman Filter for Inference of Aircraft Position and Taxi Mode from Surface Surveillance Data

“…The current taxi mode is detected by selecting the filter with the least normalized squared error, combined with a heuristic that offers resistance to rapid mode jumping. The multi-modal formulation has been previously used in areas such as computer vision, 6 but as far as we know, it has not been used before for aircraft position tracking. In this paper, we present the results from our version of the multi-modal filter, and compare it to raw data and a single-mode filter.…”

Incremental Human Motion Map system and human walking behavior representation in indoor environment

“…On environment mapping, many researchers are working on Simultaneous Localization and Mapping (SLAM) for solving problem on navigating mobile robots moving in unknown environments, and significant results are published in last decade [1]. On acquisition of human motion, Mozos and Kurazume proposed a human detection method by using multiple 2D Laser Range Finder (LRF) [2] Bellotto and Hu proposed an algorithm for human tracking by fusing sensing information from a camera and a LRF sensor installed on a mobile robot [3]. Matsumoto et al proposed an algorithm to estimate posture of human by incorporating multi-LRFs measurements of human's head and body [4].…”