Abstract:Matching objects across multiple cameras with non-overlapping views is a necessary but difficult task in the wide area video surveillance. Owing to the lack of spatio-temporal information, only the visual information can be used in some scenarios, especially when the cameras are widely separated. This paper proposes a novel framework based on multi-feature fusion and incremental learning to match the objects across disjoint views in the absence of space–time cues. We first develop a competitive major feature histogram fusion representation (CMFH11CMFH is the abbreviation of Competitive Major Feature Histogram fusion representation.) to formulate the appearance model for characterizing the potentially matching objects. The appearances of the objects can change over time and hence the models should be continuously updated. We then adopt an improved incremental general multicategory support vector machine algorithm (IGMSVM22IGMSVM is the abbreviation of Incremental General Multicategory Support Vector Machine learning algorithm.) to update the appearance models online and match the objects based on a classification method. Only a small amount of samples are needed for building an accurate classification model in our method. Several tests are performed on CAVIAR, ISCAPS and VIPeR databases where the objects change significantly due to variations in the viewpoint, illumination and poses. Experimental results demonstrate the advantages of the proposed methodology in terms of computational efficiency, computation storage, and matching accuracy over that of other state-of-the-art classification-based matching approaches. The system developed in this research can be used in real-time video surveillance applications.

Multi-Agent Fusion With Different Limited Fields-of-View

Principled information fusion for multi-view multi-agent surveillance systems

Multiagent Information Fusion and Cooperative Control in Target Search

Multi-Sensor Multi-Object Tracking with Different Fields-of-View Using the LMB Filter.

Online Multi-Object Tracking from A Bird's-Eye View by Fusion of Millimeter-Wave Radar and Vision

Distributed Fusion of Highly Maneuvering Multi-Target under Limited Field of View Sensors

Multi-Sensor Image Fusion: A Survey of the State of the Art

Exploiting Structure for Optimal Multi-Agent Bayesian Decentralized Estimation

Multi-Agent Cooperative Target Search

Video object matching across multiple non-overlapping camera views based on multi-feature fusion and incremental learning.

Distributed Multi-sensor Multi-view Fusion Based on Generalized Covariance Intersection.

Vision-based multi-agent cooperative target search

Distributed Fusion with Multi-Bernoulli Filter based on Generalized Covariance Intersection

Computationally Efficient Distributed Multi-sensor Fusion with Multi-Bernoulli Filter

Distributed Multi-Sensor Fusion of PHD Filters With Different Sensor Fields of View

S Y ] 1 6 M ar 2 01 9 Distributed Multi-sensor Multiview Fusion based on Generalized Covariance Intersection

Distributed Sensor Fusion for RFS Density with Consideration of Limited Sensing Ability

An Adaptive Fusion Strategy for Distributed Information Estimation Over Cooperative Multi-Agent Networks.

Average Marginal Density Based Distributed Multichannel Fusion For Multi-Target Tracking

Distributed Multi-View Multi-Target Tracking Based on CPHD Filtering

Multi-Sensor Multi-Frame Detection Based on Posterior Probability Density Fusion.