جستجوی مقالات مرتبط با کلیدواژه "tracking" در نشریات گروه "برق"

Tracking filters are extensively used within object tracking systems in order to provide consecutive smooth estimations of position and velocity of the object with minimum error. Namely, Kalman filter and its numerous variants are widely known as simple yet effective linear tracking filters in many diverse applications. In this paper, an effective method is proposed for designing and implementation of a Kalman filter in an object tracking application. The considered tracking application implies the capability to produce a smooth and reliable output stream by the tracking filter, even in presence of different disturbing types of noise, including background or spontaneous noises, as well as disturbances with continues or discrete nature. The presented method includes a fixed-point implementation of the Kalman filter on FPGA, which targets the joint estimation of position-velocity pair of an intended object in heavy presence of noise. The execution speed of the Kalman algorithm is drastically enhanced in the proposed implementation. This enhancement is attained by emphasis on hardware implementation of every single computational block on the one hand, and through appropriate parallelization and pipelining of independent tasks within the Kalman process on the other hand. After designing the filter parameters with respect to the requirements of a given tracking problem, a floating-point model and a fixed-point hardware model of the filter are implemented using MATLAB and Xilinx System Generator, respectively. In order to evaluate the performance of the filter under realistic circumstances, a set of appropriately defined scenarios are carried out. The simulations are carefully designed in order to represent the extremely harsh scenarios in which the input measurements to the filter are deeply polluted by different kinds of noises. In each simulation the position-velocity data corresponding to a moving object is generated according to an appropriate model, quantized, and contaminated by noise and fed into the filter. Performances of the Kalman filter in software version (i.e. the floating point replica) and hardware version (i.e. the fixed-point replica) are quantitatively compared in the designed scenario. Our comparison employs NMSE and maximum error values as quantitative measures, verifying the competency of our proposed fixed-point hardware implementation. The results of our work show that, with adequate selection word length, the implemented filter is fast and efficient; it confines the algorithm execution time to 50 clock pulses, i.e. about 0.4 µs when a 125 MHz clock is used. It is also verified that our implementation reduces the position and velocity estimation errors by 11 dB and 1.2 dB, respectively. The implemented filter also confines the absolute values of maximum error in position and velocity to 10 meter and 0.7 meter/sec. in the considered scenario, which is almost resembles the performance of its floating point counterpart. The presented Kalman filter is finally implemented on Zc706 evaluation board and the amount of utilized hardware resource (FFs, LUTs, DSP48, etc.) are reported as well as the estimated power consumption of the implemented design. The paper is concluded through comparison of the proposed design with some recent works which confirms the efficacy of the presented implementation.

Detecting and tracking of moving objects is an important task in analyzing videos. In this paper, we propose a new method for tracking several concurrent moving objects of fixed camera. In the proposed method, at each stage, the location of moving objects in front of camera view is obtained information between two current and previous frames. In each step, Sift’s edge points is obtained based on previous frame and to get the correspondence of these feature points by the use of KLT feature point correspondence algorithm on the current frame. Then having correspondent feature points between two sequence frames, we would estimate the distance by eliminating partial or fixed moving feature points related to moving objects. The classification of labeled features as moving objects is done using DBSCAN clustering algorithm into different clusters. By this method and on each moment, the situation of all existing moving objects in camera view which has got by one by one correspondence between these objects, is determined. The obtained results of the proposed method shows a high degree of accuracy and acceptable consuming time to track moving objects.

In ground moving target tracking missions by UAVs, there are many reasons, such as running out of battery or fuel, to handover the mission to another UAV. In this paper we present an online handover algorithm with low probability of losing target while trying to maximize regional coverage along the takeover trajectory. The UAV that takes over the target predicts the target path using the estimated motion model and the last position of the target that is provided by the other UAV. Afterwards, the uncertainty surrounding each position of the target over the predicted path is computed. The position of the target in which the position uncertainty and the UAV's field of view have minimum difference, is the best place for retrieving the target. Among the possible paths to the retrieval point, the path that provides the most coverage is the best retrieval path. To determine this path, we have proposed an algorithm based on a “forward selection” strategy. In this algorithm, the uncovered and important positions of the region are clustered based on the distance from each other. Then, the centers of these clusters are added to the previous retrieved path recursively based on their effectiveness on the coverage value if the time of traveling on the new path is smaller than the time required for the target to reach the retrieval point. In each execution of the algorithm, using a “backward elimination” strategy, the coverage value of the retrieval path after recursive combination of each two adjacent clusters is checked to avoid from falling in local optima. We have simulated the proposed method and compared it with the optimal offline calculated path which shows the success of the proposed approach in retrieving the target and maximizing the regional coverage.

Navigation and localization are two key issues for many modern mobile applications. Navigation data can be gathered either passively (i.e. by means of INS) or actively (e.g. GPS, Loran). the passive navigation data Usually degrade through the time. On the other hand, the active methods are prune to interference and jamming. In this paper, the passive navigation method is aided by the measured angle of arrival from emitting objects (i.e. radars) at known locations. The results show that in the proposed method in contradiction to many other passive methods, while maintaining the passiveness of the system, using AoA measurement, the quality of navigation data is preserved or even improved through the time.

In this paper Intelligent image processing-based rapid method for detecting and tracking moving vehicles at intersections is proposed. In the detection part, the Gaussian mixture model used to obtain the moving parts. Then, the targets were detected using HOG features extracted from training images, Adaboost Cascade Classifier and the trained SVM.
At the tracking part, a number of key points on the image of the vehicle were identified at first. The center of mass of the object and the edges were used to obtain these key points because these points are primarily important and more common in tracking rigid bodies. Then, these points were tracked in consecutive frames using definitive adaptive procedures. Also, the Kalman filter was used to estimate new locations when the detector was not able to detect the targets.
The major advantage of this method compared to previous methods, is resistance against vehicle's overlapping and changes in Illuminations,so that the detection accuracy is 90.80% on overloaded traffic scenes and 88.75% accuracy on the tracking vehicles.

The population of elderly people has growing trend in the developed or developing countries. Since the elderly are mainly associated with disability, this group of people exposed to dangerous events such as falling down. It is therefore needed to take care of these people against dangerous events. Intelligent video monitoring is a approach that may be able to give quick notification to the caregivers. For this purpose, in this paper,an approach using a novel tracking method based on modified contour algorithm is presented. The new method is able to conduct tracking and falling down detection in a realistic conditions in the presence of multiple motions. Simulations indicate that the proposed algorithm is able to identify the fall-down event with high accuracy and speed.

People re-identification is one of the most important and fundamental processes in video surveillance systems. The accuracy and efficiency of this task influence the effectiveness of the subsequent processes. Event detection and behavior analysis are instances of such subsequent processes that are classified in semantic levels. In people re-identification, having an image or video of an individual in a specific camera, we want to infer whether the person has been seen previously in the other cameras. Various changes in people’s appearance, recorded by different cameras, can lead to many difficulties in people re-identification. These changes may occur due to hardware limitations such as low-resolution, different color responses of cameras, lighting changes in different camera locations and person’s angle relative to the camera. In this paper we will present a review on the existing approaches in people re-identification, then, a new classification of these works based on their functionalities is provided. Our studies shows that the problem of changing the camera angles has not been seriously considered in the existing approaches and future research could provide models that able to resolve these problems.

In this paper، a method to track the maximum power of solar panels based on fuzzy logic is presented. The proposed method is based on the relationship between radiation intensity and the voltage of maximum power operating point. With this relationship، at any time by measuring the light intensity، voltage can be calculated at the maximum power point by using fuzzy approximation function. In order to verify the proposed method، simulation results are presented.

LEO satellites orbit very fast respect to the Earth stations, so the tracking of these satellites is important. But due to some challenges, such as structural deformations caused either by loads or by the temperature gradient, atmospheric distortion, or azimuth track unevenness, the position of beam is slightly different from the antenna position as measured by the encoders. Thus, the monopulse technique as the most accurate type of Auto-tracking techniques is used and analyzed. Also, the reduction of beam error and increment of received power versus signal to noise ratio are investigated. Finally, estimation parameters such as variance, mean and RMSE of monopulse technique versus signal to noise ratio are investigated.