Co-operative Multi-target Tracking and Classification

2008

Non-intrusive video-detection for traffic flow observation and surveillance is the primary alternative to conventional inductive loop detectors. Video Image Detection Systems (VIDS) can derive traffic parameters by means of image processing and pattern recognition methods. Existing VIDS emulate the inductive loops. We propose a trajectory based recognition algorithm to expand the common approach and to obtain new types of information (e.g. queue length or erratic movements).

Data association is an essential component of the human detection and tracking system. The majority of the existing methods, such as Bi-partite matching and GMCP methods are incorporated the limited-temporal-locality of the sequence into data association problem. GMCP tracker is considered as an important complete representation of the tracking problem, where all pair wise relationships between the detections in temporal span of a video is considered and makes the input to the data association as a complete Bi-partite graph. In Bi-partite graph a track of a person will form a clique (a subgraph in which all the nodes are connected to each other). A cost is assigned to each clique and it maximizes the score function, which is selected as the best clique (track), but it is sub-optimal. GMCP tracker does not follow the joint optimization for all the tracks simultaneously and finds the tracks one by one which makes difficulties caused by cluttered background, and crowded scenes to detect and tracking Tracking-by-detection methods are used to track multiple targets with unified handling of complex scenarios, where current detection responses are linked to the previous trajectories. By adding the standard Hungarian algorithm, dummy nodes to each trajectory to allow nodes to temporally disappear and solve the data association implicitly in a global manner even though it is formulated between two consecutive frames. If a trajectory fails to find its matching detection, it is linked to its corresponding dummy nodes until its emergence of matching detection. The source nodes are also incorporated into the account of new targets. The dummy nodes tend to accumulate in fake or disappeared trajectories while they occasionally appear in real trajectories and improve detection inevitable failures, which include the miss detection, the false detection and the occlusion, where an object is partially or fully invisible because of the limited camera view. Extended hybrid Hungarian algorithm is relatively better when compared with GMCP and Hybrid Hungarian algorithm in accuracy. Experiments show that the proposed method makes significant improvement in tracking and detection of different length of videos, specifically with short length videos.

2007 Ieee International Symposium on Industrial Electronics, Proceedings, Vols 1-8, 2007

In this paper one of the most important solutions in position estimation is used in conjunction with a data association algorithm in order to achieve a multi-tracking application. A Kalman Filter is extended and adapted in order to track the position and speed of a variable number of objects in an unstructured and complex environment. Both the developed algorithms and the results obtained with their real-time execution implementation in the mentioned application are described, and interesting conclusions extracted from these experiments are remarked in the paper. Finally, tracking results of the proposed algorithm are compared with another multi-object estimator based on a Particle Filter previously developed by the authors. I.

In this paper, we study the problem of joint tracking and classification of several targets at the same time. Targets are considered to be known and sufficiently separated so that they cannot be confused. Our goal is to propose a full methodology that is robust to missing information. The classical probabilistic approach with Bayesian tools is improved with belief functions. A simulation concerning the identification of go fast boats in a piracy problem shows that our approach improves previous results.

IEE Seminar on Target Tracking: Algorithms and Applications, 2006

This paper presents a particle filtering algorithm for multiple object tracking. The proposed particle filter (PF) embeds a data association technique based on the joint probabilistic data association (JPDA) which handles the uncertainty of the measurement origin. I. Introduction Tracking a group of targets in video sequences has many surveillance applications. It has been used for security monitoring (1), (2),

2012

Multiple target tracking that integrates target model estimation and data association steps is described. The integration allows successive refinement of the models while reducing the uncertainty in data association. Each target is described by "weak" models of kinematics, shape and appearance. The target models are refined in a two-stage process: imagebased tracklets of high purity and accuracy are generated, and geospatial tracks are extended from these tracklets. During each stage of tracking, observation data of reduced uncertainties are associated with the refined tracks in a probabilistic manner. We describe our approach in the context of a real time system that has been tested and evaluated for vehicle and human tracking in sparse, medium, and dense clutter using aerial EO/IR video.

2017

In this paper, we consider multi-object target tracking using video reference datasets. Our objective is detection of the target using a novel adaboost and Gentle Boost method in order to track the subjects from reference data sets. Multi-target tracking is still challenging topic which is used to find the same object across different camera views and also used to find the location and sizes of different object at different places. Furthermore extensive performance analysis of the three main parts demonstrates usefulness of multi object tracking. We carried out experiment to analyze discriminative power of nine features (HSV, LBP, HOG extracted on body, torso and legs) used in the appearance model for multicam dataset. For each features the RMSE and PSNR obtained.

Lecture Notes in Computer Science, 2006

In this paper, we propose a maximum a posteriori formulation to the multiple target tracking problem. We adopt a graph representation for storing the detected regions as well as their association over time. The multiple target tracking problem is formulated as a multiple paths search in the graph. Due to the noisy foreground segmentation, an object may be represented by several foreground regions and one foreground region may corresponds to multiple objects. We introduce merge, split and mean shift operations that add new hypothesis to the measurement graph in order to be able to aggregate, split detected blobs or re-acquire objects that have not been detected during stop-and-gomotion. To make full use of the visual observations, we consider both motion and appearance likelihood. Experiments have been conducted on both indoor and outdoor data sets, and a comparison has been carried to assess the contribution of the new tracker.

2010 13th International Conference on Information Fusion, 2010

With the evolution and fusion of technologies from sensor networks and embedded cameras, smart camera networks are emerging as useful and powerful systems. Wireless networks, however, introduce new constraints of limited bandwidth, computation, and power. Existing camera network approaches for target tracking either utilize target handover mechanisms between cameras, or combine results from 2D trackers into 3D target state for continuous tracking. Such approaches suffer from the drawbacks associated with 2D tracking, such as scale selection, target rotation, and occlusion. In this paper, we present an approach for tracking multiple targets in 3D space using a wireless network of smart cameras. In our approach, we use multiview histograms in different feature-spaces to characterize targets in 3D space. We employ color and texture as the visual features to model targets. The visual features from each camera, along with the target models are used in a probabilistic tracker to estimate the target state. We demonstrate the effectiveness of our proposed tracker with results tracking people using a camera network deployed in a building.

16th IPPR Conference on CVGIP, 2003

Moving object tracking has been an important research topic in numerous applications. Most papers adopted estimation-based approach to achieve moving object tracking. In this paper, an effective tracking approach using contour and texture features is proposed. Detail analysis of the proposed features for tracking is elaborated. A tracking scheme by classifying the features with backpropagation neural network is invented. The proposed approach is verified through a set of experimental image sequences. 1310 moving objects, including 486 cars and 824 humans, can be effectively classified with 99.8% recognition rate.

ICTACT Journal on Image and Video Processing, 2017

2018

Most multiple object tracking algorithms relying on a single view have failed to follow the trajectories of targets when they have been completely hidden by obstacles. In this paper, we introduce a novel method of collaborative tracking in a synchronized overlapping cameras network. We propose an efficient target association method between cameras based on the tracking results of each target on each view. Our framework naturally handles obstacle occlusions and mutual target occlusions. We implemented our multiple object tracking algorithm by Decision Making algorithm [30] on each view. The tracking outcomes on each camera are collected and associated into targets. The feedback from the central association helps the individual cameras in tracking hidden targets, even in the case of complete occlusion. We use the standard MOT metric to validate our method. The experimental results on each view show that the multiple view tracking system outperforms the single view ones. The source cod...

Pattern Recognition Letters, 2010

Visual detection and tracking are interdisciplinary tasks which are oriented at estimating the state of one or multiple moving objects in a video sequence. This is one of the first tasks in processing video systems which try to describe human behaviour in different contexts, such as video-surveillance, sport technique analysis. This work presents a multiple object tracking system which properly hybridizes particle filters and memetic algorithms to produce a more reliable and efficient tracking algorithm. The system has been tested on synthetic and real image sequences, with the aim of describing their performance for different levels of noise, occlusions, a variable number of objects, etc. Experimental results demonstrate that the proposed system accurately tracks multiple objects in the scene, by grouping and ungrouping them when necessary, while keeping their identities during the sequence of images. Moreover, the performance of the proposed system is not strongly affected by the increase in the number of objects, maintaining computational load and precision in proper balance.

IEEE Signal Processing Magazine, 2002

We outline a framework for collaborative signal processing in distributed sensor networks. The ideas are presented in the context of tracking multiple moving objects in a sensor field. The key steps involved in the tracking procedure include event detection, target classification, and estimation and prediction of target location. Algorithms for various tasks are discussed with an emphasis on classification. Results based on experiments with real data are reported which provide useful insights into the essential nature of the problems. Issues, challenges and directions for future research are identified.