Globally optimal target tracking in real time using max-flow network

Abstract

We propose a general framework for multiple target tracking across multiple cameras using max-flow networks. The framework integrates target detection, tracking, and classification from each camera and obtains the cross-camera trajectory of each target. The global data association problem is formed as a maximum a posteriori (MAP) problem and represented by a flow network. Similarities of time, location, size, and appearance (classification and color histogram) of the target across cameras are provided as inputs to the network and the target's optimal cross-camera trajectory is found using the max-flow algorithm. The implemented system is designed for real-time process with high-resolution videos (10MB per frame). The framework is validated on high resolution camera networks with both overlapping and non-overlapping fields of view in urban scenes.

Distributed data association in smart camera network via dual decomposition

“…MCF was originally proposed for multi-targets tracking in a single camera view [21], and recently has been used for data association across disjoint camera views [16]. In our implementation, the cost-flow network is defined as that in Fig.3(b), augmented with a source and a sink node, both connecting to all observation nodes in the network.…”

“…One of the fundamental prerequisites for achieving these goals is the correct reconstruction of camera-to-camera trajectory of each object, or equivalently, grouping observations originated from the same object into a single track, which may be generated by different cameras at different time instants. This problem is often referred to as data association in camera networks [1,12], trajectory recovery [13], or camera-to-camera tracking [14][15][16][17][18]. In this paper, we assume that the detection and tracking problem within a single camera view has been solved, and we call the collection of quantities summarizing the features of tracked object as a virtual "observation", see Fig.2 as an example.…”

Distributed Bayesian Inference for Consistent Labeling of Tracked Objects in Nonoverlapping Camera Networks

“…The problem of consistent labeling in camera networks using both appearance and spatiotemporal cues has been widely investigated, under the name of data association [1,10], trajectory recovery [11], or camera-to-camera tracking [12][13][14][15][16][17]. Some authors try to solve the problem by optimally partitioning the set of observations collected by the camera networks into several disjoint subsets, such that the observations in each subset are believed to come from a single object.…”

“…The topological assumption imposes strong constraints on the movements of objects in the networks, in that any object, if not newly appeared, must be presented in the FOV of one of the current camera's neighbors before it arrives to the current camera. It holds true in many scenarios of interest and has been adopted in most of the works about tracking in nonoverlapping camera networks [1,2,[10][11][12][13][14][15][16][17].…”

“…Some authors try to solve the problem by optimally partitioning the set of observations collected by the camera networks into several disjoint subsets, such that the observations in each subset are believed to come from a single object. The difficulty caused by the exponential growth of the partition space is tackled by making appropriate independence assumptions and leveraging efficient optimization algorithms, such as Markov Chain Monte Carlo [11, 14], Max-Flow Network [15], or Multiple Hypothesis Tracking [16, 17]. On the other hand, it is more attractive to treat the problem in a Bayesian framework [1, 10], in which each observation is assigned with a labeling variable, and the posterior distribution of the labeling variable is inferred, conditioned on all evidence made in whole networks.…”