Sharif Digital Repository

While Probabilistic MHT (PMHT) highly reduces computational burden by avoiding enumeration of all possible data association events, its implementation still suffers from intractable manifold integrals. MCMC particle filters, rendering the task of PMHT implementation, found it difficult to face with both data association uncertainty and violations of the presumptions of no more than one measurement could be assigned to a target at each time step. Coping with these difficulties, we suggest a modified Gibbs sampling approach that utilizes current measurement information efficiently to estimate PMHT parameters recursively. We use scenarios of Bearing-Only multi-target tracking to compare our... 

We study the problem of power efficient tracking interval management for distributed target tracking wireless sensor networks (WSNs). We first analyze the performance of a distributed target tracking network with one moving object, using a quantitative mathematical analysis. We show that previously proposed algorithms are efficient only for constant average velocity objects; however, they do not ensure an optimal performance for moving objects with acceleration. Towards an optimal performance, first, we derive a closed-form mathematical expression for the estimation of the minimal achievable power consumption by an optimal adaptive tracking interval management algorithm. This can be used as... 

This paper investigates maneuvering-target tracking problem based on a multisensor system and interacting multiple model (IMM). The estimation is performed by a novel particle filter (PF) with a capability to deal with the state-dependent noises and interference of the sensors’ coverage environment. An adaptive sensor selection algorithm, where some sensors are selected in each stage based on the signal-to-interference pulse noise ratio (SINR) and participate in the state estimation, is proposed. To deal with the effect of interference, we focus on designing and implementing the sensor selection algorithm, where a multisensor system with nonuniform arrays is derived by solving a convex... 

This paper presents a new statistical based clean algorithm for multiple target detection in noisy environments. For this purpose the variance and average of the ambiguity function as the output of matched filter will be computed. This information used for separating the real targets from the false ones, in noisy environment and heavy clutter up to - 13 dB. © 2006 IEEE  

In this paper the accuracy of measurement with respect to some parameters like baseline variance and DOA estimation and bistatic RCS is considered. The Radar designer can find the appropriate parameters to having the minimum error in target localization and maximum coverage range. © 2006 IEEE  

A new fast and very accurate algorithm for target detection in PCL (Passive Coherent Location) radars is presented. This algorithm in noisy environment that SNR is low as -45dB operates, with an error less than 20 percent. Presented algorithm is capable for target detection by few samples of signals and obtains real-time processing in passive radars. © 2006 IEEE  

A new approach in multiple target detection in PCL (Passive Coherent Location) radars based on TV and Radio ambiguity function processing is presented. Fast computation and high Accuracy are the presented algorithm capabilities. Presented algorithm is a new and simple method which could provide perfect detection in noisy environment up to SNR= -30 dB  

Registration is the most consequential topic to be dealt with in a multi-sensor tracking system. On the other hand, providing satisfactory target acceleration estimation would enhance the performance of a ground-based air defense system encountering a maneuvering target. The main novelty of the present work is addressing a new scheme to solve the registration problem in a distributed network along with estimating accurate target acceleration, simultaneously. The details of coping with three common kinds of measurement, attitude, and location biases are explored concentrating on the effects of attitude bias as the main error source from the practical viewpoint. A modified iterated extended... 

The Gaussian mixture probability hypothesis density (GM-PHD) filter is a closed form approximation of the multi-target Bayes filter which can overcome most multitarget tracking problems. Limited field of view, decreasing cost of cameras, and advances of using multi-camera induce us to use large-scale camera networks. In this paper, a multihuman tracking framework using the PHD filter in a distributed camera network is proposed. Each camera tracks objects locally with PHD filter and a track-after-detect scheme and its estimates of targets are sent to neighboring nodes. Then each camera fuses its local estimates with it's neighbors. The proposed method is evaluated on the public PETS2009... 

In this paper the problem of cooperative tracking of multiple targets for multi-agent systems is investigated. Defining problem of target tracking as gathering maximum rewards associated with the targets, an optimization-based algorithm is presented. A cooperative receding horizon controller for reaching moving targets is proposed. Agents are controlled by adjusting their headings toward the moving targets. A notable advantage of the proposed approach is the estimations of target's motion and planning the agent's heading based on them. Simulation results are provided to verify the efficiency of the proposed method  

While visual target tracking is one of the noteworthy and the most active research areas in computer vision and machine learning, many challenges are still unresolved. In this paper, an adaptive generic target tracker is proposed that includes the adaptive determination of learning parameters from spatio-temporal context model, analysis of prior targets and confidence map for accurate target localization, and modified scale estimation scheme based on confidence map. According to spatio-temporal context model, the learning parameters are adaptively determined for achieving confidence map and target scale robustly. Moreover, analysis of the confidence map helps our tracker to change context... 

Visual target tracking is one of the most sought-after yet challenging research topics in computer vision. Given the ill-posed nature of the problem and its popularity in a broad range of real-world scenarios, a number of large-scale benchmark datasets have been established, on which considerable methods have been developed and demonstrated with significant progress in recent years - predominantly by recent deep learning (DL)-based methods. This survey aims to systematically investigate the current DL-based visual tracking methods, benchmark datasets, and evaluation metrics. It also extensively evaluates and analyzes the leading visual tracking methods. First, the fundamental... 

Visual target tracking is one of the most sought-after yet challenging research topics in computer vision. Given the ill-posed nature of the problem and its popularity in a broad range of real-world scenarios, a number of large-scale benchmark datasets have been established, on which considerable methods have been developed and demonstrated with significant progress in recent years - predominantly by recent deep learning (DL)-based methods. This survey aims to systematically investigate the current DL-based visual tracking methods, benchmark datasets, and evaluation metrics. It also extensively evaluates and analyzes the leading visual tracking methods. First, the fundamental... 

A new approach for tracking targets in TWS systems is introduced which is based on Hough transform. The main idea is to take target's velocity and course of motion as the main tracking quantities and calculate other kinematic parameters of the target from these two main parameters. The main criteria for associating targets to tracks in Data Association (DA) step is co linearity of targets with the history of track and Hough transform is the main tool in this step. Computer simulation results are presented to compare the performance of the suggested algorithm with the conventional tracking systems  

This paper investigates the problem of unknown input estimation such as acceleration, target class, and maneuvering target tracking using a hybrid algorithm. One of the challenges of unknown input estimation is that no effective method has been presented so far that could be applied to general cases. The available methods are ineffective when the range of variation of the unknown input parameter is large. Also, the issue of determining the system class could improve the performance of the tracking algorithms in many applications. Using the Bayesian theory, the posterior distribution functions of state and parameter could be obtained concurrently. In the proposed algorithm, Liu and West and... 

Using the Fractal Geometry in Signal processing has been extended nowadays [1]. They have found several applications in signal detection and recognitions. They use the chaotic feature of the noise and clutter and try to distinguish between the noise, clutter and the desired target [2] [3]. Recent works show that lots of clutters like sea and ground clutters have fractal behavior so this kind of approach to the signal detection has been extended these days[3][4]. In this paper we have used the Box-Counts of a signal rather than the Fractal dimension of a signal as will be defined later in the text. By applying the new defined concept we have developed different methods of signal detection. We...