Abstract: Multiple classifier systems combine several individual classifiers to deliver a final classification decision. However, an increasingly controversial question is whether such systems can outperform the single best classifier, and if so, what form of multiple classifiers system yields the most significant benefit. Also, multi-target tracking detection using multiple sensors is an important research field in mobile techniques and military applications. In this paper, several multiple classifiers systems are evaluated in terms of their ability to predict a system’s failure or success for multi-sensor target tracking tasks. The Bristol Eden project dataset is utilised for this task. Experimental and simulation results show that the human activity identification system can fulfil requirements of target tracking due to improved sensors classification performances with multiple classifier systems constructed using boosting achieving higher accuracy rates.
Abstract: Kernelized Correlation Filter (KCF) based trackers have gained a lot of attention recently because of their accuracy and fast calculation speed. However, this algorithm is not robust in cases where the object is lost by a sudden change of direction, being obscured or going out of view. In order to improve KCF performance in long-term tracking, this paper proposes an anomaly detection method for target loss warning by analyzing the response map of each frame, and a classification algorithm for reliable target re-locating mechanism by using Random fern. Being tested with Visual Tracker Benchmark and Visual Object Tracking datasets, the experimental results indicated that the precision and success rate of the proposed algorithm were 2.92 and 2.61 times higher than that of the original KCF algorithm, respectively. Moreover, the proposed tracker handles occlusion better than many state-of-the-art long-term tracking methods while running at 60 frames per second.
Abstract: Shear elastic modulus of skeletal muscles can be
obtained by shear wave elastography (SWE) and has been
linearly related to muscle force. However, SWE is currently
implemented using array probes. Price and volumes of these probes
and their driving equipment prevent SWE from being used in
wearable human-machine interfaces (HMI). Moreover, beamforming
processing for array probes reduces the real-time performance. To
achieve SWE by wearable HMIs, a customized three-element probe
is adopted in this work, with one element for acoustic radiation
force generation and the others for shear wave tracking. In-phase
quadrature demodulation and 2D autocorrelation are adopted to
estimate velocities of tissues on the sound beams of the latter two
elements. Shear wave speeds are calculated by phase shift between
the tissue velocities. Three agar phantoms with different elasticities
were made by changing the weights of agar. Values of the shear
elastic modulus of the phantoms were measured as 8.98, 23.06 and
36.74 kPa at a depth of 7.5 mm respectively. This work verifies the
feasibility of measuring shear elastic modulus by wearable devices.
Abstract: Despite being one of the most significant economic contributors of the country, Canada’s construction industry is lagging behind other sectors when it comes to labor productivity improvements. The construction industry is very collaborative as a general contractor, will hire trade contractors to perform most of a project’s work; meaning low productivity from one contractor can have a domino effect on the shared success of a project. To address this issue and encourage trade contractors to improve their productivity tracking methods, an investigative study was done on the productivity views and tracking methods of various trade contractors. Additionally, an in-depth review was done on four standard tracking methods used in the construction industry: cost codes, benchmarking, the job productivity measurement (JPM) standard, and WorkFace Planning (WFP). The four tracking methods were used as a baseline in comparing the trade contractors’ responses, determining gaps within their current tracking methods, and for making improvement recommendations. 15 interviews were conducted with different trades to analyze how contractors value productivity. The results of these analyses indicated that there seem to be gaps within the construction industry when it comes to an understanding of the purpose and value in productivity tracking. The trade contractors also shared their current productivity tracking systems; which were then compared to the four standard tracking methods used in the construction industry. Gaps were identified in their various tracking methods and using a framework; recommendations were made based on the type of trade on how to improve how they track productivity.
Abstract: Human pose estimation and tracking are to accurately identify and locate the positions of human joints in the video. It is a computer vision task which is of great significance for human motion recognition, behavior understanding and scene analysis. There has been remarkable progress on human pose estimation in recent years. However, more researches are needed for human pose tracking especially for online tracking. In this paper, a framework, called PoseSRPN, is proposed for online single-person pose estimation and tracking. We use Siamese network attaching a pose estimation branch to incorporate Single-person Pose Tracking (SPT) and Visual Object Tracking (VOT) into one framework. The pose estimation branch has a simple network structure that replaces the complex upsampling and convolution network structure with deconvolution. By augmenting the loss of fully convolutional Siamese network with the pose estimation task, pose estimation and tracking can be trained in one stage. Once trained, PoseSRPN only relies on a single bounding box initialization and producing human joints location. The experimental results show that while maintaining the good accuracy of pose estimation on COCO and PoseTrack datasets, the proposed method achieves a speed of 59 frame/s, which is superior to other pose tracking frameworks.
Abstract: In this paper we present the design and the implementation of a target tracking system where the target is set to be a moving person in a video sequence. The system can be applied easily as a vision system for mobile robot. The system is composed of two major parts the first is the detection of the person in the video frame using the SVM learning machine based on the “HOG” descriptors. The second part is the tracking of a moving person it’s done by using a combination of the Kalman filter and a modified version of the Camshift tracking algorithm by adding the target motion feature to the color feature, the experimental results had shown that the new algorithm had overcame the traditional Camshift algorithm in robustness and in case of occlusion.
Abstract: Tracking sports players is a widely challenging
scenario, specially in single-feed videos recorded in tight courts,
where cluttering and occlusions cannot be avoided. This paper
presents an analysis of several geometric and semantic visual features
to detect and track basketball players. An ablation study is carried
out and then used to remark that a robust tracker can be built with
Deep Learning features, without the need of extracting contextual
ones, such as proximity or color similarity, nor applying camera
stabilization techniques. The presented tracker consists of: (1) a
detection step, which uses a pretrained deep learning model to
estimate the players pose, followed by (2) a tracking step, which
leverages pose and semantic information from the output of a
convolutional layer in a VGG network. Its performance is analyzed
in terms of MOTA over a basketball dataset with more than 10k
instances.
Abstract: Falls are one of the major causes of injury and death
among elderly people aged 65 and above. A support system to
identify such kind of abnormal activities have become extremely
important with the increase in ageing population. Pose estimation
is a challenging task and to add more to this, it is even more
challenging when pose estimations are performed on challenging
poses that may occur during fall. Location of the body provides a
clue where the person is at the time of fall. This paper presents
a vision-based tracking strategy where available joints are grouped
into three different feature points depending upon the section they are
located in the body. The three feature points derived from different
joints combinations represents the upper region or head region,
mid-region or torso and lower region or leg region. Tracking is always
challenging when a motion is involved. Hence the idea is to locate
the regions in the body in every frame and consider it as the tracking
strategy. Grouping these joints can be beneficial to achieve a stable
region for tracking. The location of the body parts provides a crucial
information to distinguish normal activities from falls.
Abstract: Radiologists routinely make follow-up imaging recommendations, usually based on established clinical practice guidelines, such as the Fleischner Society guidelines for managing lung nodules. In order to ensure optimal care, it is important to make guideline-compliant recommendations, and also for patients to follow-up on these imaging recommendations in a timely manner. However, determining such compliance rates after a specific finding has been observed usually requires many time-consuming manual steps. To address some of these limitations with current approaches, in this paper we discuss a methodology to automatically detect finding-specific follow-up recommendations from radiology reports and create a visualization for relevant subsequent exams showing the modality transitions. Nearly 5% of patients who had a lung related follow-up recommendation continued to have at least eight subsequent outpatient CT exams during a seven year period following the recommendation. Radiologist and section chiefs can use the proposed tool to better understand how a specific patient population is being managed, identify possible deviations from established guideline recommendations and have a patient-specific graphical representation of the imaging pathways for an abstract view of the overall treatment path thus far.
Abstract: This paper presents a real time video surveillance system which is capable of tracking multiple real time objects using Polar Vector Median (PVM) and Block Coding Modes (BCM) with Global Motion Compensation (GMC). This strategy works in the packed area and furthermore utilizes the movement vectors and BCM from the compressed bit stream to perform real time object tracking. We propose to do this in view of the neighboring Motion Vectors (MVs) using a method called PVM. Since GM adds to the object’s native motion, for accurate tracking, it is important to remove GM from the MV field prior to further processing. The proposed method is tested on a number of standard sequences and the results show its advantages over some of the current modern methods.
Abstract: Hand gesture recognition is a technique used to locate, detect, and recognize a hand gesture. Detection and recognition are concepts of Artificial Intelligence (AI). AI concepts are applicable in Human Computer Interaction (HCI), Expert systems (ES), etc. Hand gesture recognition can be used in sign language interpretation. Sign language is a visual communication tool. This tool is used mostly by deaf societies and those with speech disorder. Communication barriers exist when societies with speech disorder interact with others. This research aims to build a hand recognition system for Lesotho’s Sesotho and English language interpretation. The system will help to bridge the communication problems encountered by the mentioned societies. The system has various processing modules. The modules consist of a hand detection engine, image processing engine, feature extraction, and sign recognition. Detection is a process of identifying an object. The proposed system uses Canny pruning Haar and Haarcascade detection algorithms. Canny pruning implements the Canny edge detection. This is an optimal image processing algorithm. It is used to detect edges of an object. The system employs a skin detection algorithm. The skin detection performs background subtraction, computes the convex hull, and the centroid to assist in the detection process. Recognition is a process of gesture classification. Template matching classifies each hand gesture in real-time. The system was tested using various experiments. The results obtained show that time, distance, and light are factors that affect the rate of detection and ultimately recognition. Detection rate is directly proportional to the distance of the hand from the camera. Different lighting conditions were considered. The more the light intensity, the faster the detection rate. Based on the results obtained from this research, the applied methodologies are efficient and provide a plausible solution towards a light-weight, inexpensive system which can be used for sign language interpretation.
Abstract: In this study, a model, together with a software tool that implements it, has been developed to determine the performance ratings of employees in an organization operating in the information technology sector using the indicators obtained from employees' online study data. Weighted Sum (WS) Method and Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) method based on multidimensional decision making approach were used in the study. WS and TOPSIS methods provide multidimensional decision making (MDDM) methods that allow all dimensions to be evaluated together considering specific weights, allowing employees to objectively evaluate the problem of online performance tracking. The application of WS and TOPSIS mathematical methods, which can combine alternatives with a large number of dimensions and reach simultaneous solution, has been implemented through an online performance tracking software. In the application of WS and TOPSIS methods, objective dimension weights were calculated by using entropy information (EI) and standard deviation (SD) methods from the data obtained by employees' online performance tracking method, decision matrix was formed by using performance scores for each employee, and a single performance score was calculated for each employee. Based on the calculated performance score, employees were given a performance evaluation decision. The results of Pareto set evidence and comparative mathematical analysis validate that employees' performance preference rankings in WS and TOPSIS methods are closely related. This suggests the compatibility, applicability, and validity of the proposed method to the MDDM problems in which a large number of alternative and dimension types are taken into account. With this study, an objective, realistic, feasible and understandable mathematical method, together with a software tool that implements it has been demonstrated. This is considered to be preferable because of the subjectivity, limitations and high cost of the methods traditionally used in the measurement and performance appraisal in the information technology sector.
Abstract: A method for object tracking in motion blurred images is proposed in this article. This paper shows that object tracking could be improved with this approach. We use mean shift algorithm to track different objects as a main tracker. But, the problem is that mean shift could not track the selected object accurately in blurred scenes. So, for better tracking result, and increasing the accuracy of tracking, wavelet transform is used. We use a feature named as blur extent, which could help us to get better results in tracking. For calculating of this feature, we should use Harr wavelet. We can look at this matter from two different angles which lead to determine whether an image is blurred or not and to what extent an image is blur. In fact, this feature left an impact on the covariance matrix of mean shift algorithm and cause to better performance of tracking. This method has been concentrated mostly on motion blur parameter. transform. The results reveal the ability of our method in order to reach more accurately tracking.
Abstract: Vehicle tracking and accident recognizing are considered by many industries like insurance and vehicle rental companies. The main goal of this paper is to detect the location of a car accident by combining different methods. The methods, which are considered in this paper, are Global Navigation Satellite Systems/Inertial Measurement Units (GNSS/IMU)-based navigation and vehicle accident detection algorithms. They are expressed by a set of raw measurements, which are obtained from a designed integrator black box using GNSS and inertial sensors. Another concern of this paper is the definition of accident detection algorithm based on its jerk to identify the position of that accident. In fact, the results convinced us that, even in GNSS blockage areas, the position of the accident could be detected by GNSS/INS integration with 50% improvement compared to GNSS stand alone.
Abstract: In the current work, adaptive extended Kalman filter (AEKF) is presented for solution of ground radar based ballistic missile (BM) tracking problem in re-entry phase with unknown ballistic coefficient. The estimation of trajectory of any BM in re-entry phase is extremely difficult, because of highly non-linear motion of BM. The estimation accuracy of AEKF has been tested for a typical test target tracking problem adopted from literature. Further, the approach of AEKF is compared with extended Kalman filter (EKF). The simulation result indicates the superiority of the AEKF in solving joint parameter and state estimation problems.
Abstract: The Lens-Lens Beam Generator (LLBG) is a Fresnel-based optical concentrating technique which provides flexibility in selecting the solar receiver location compared to conventional techniques through generating a powerful concentrated collimated solar beam. In order to achieve that, two successive lenses are used and followed by a flat mirror. Hence the generated beam emerging from the LLBG has a high power flux which impinges on the target receiver, it is important to determine the precision of the system output. In this present work, mathematical investigation of different parameters affecting the precision of the output beam is carried out. These parameters include: Deflection in sun-facing lens and its holding arm, delay in updating the solar tracking system, and the flat mirror surface flatness. Moreover, relationships that describe the power lost due to the effect of each parameter are derived in this study.
Abstract: In this paper, a scalable augmented reality framework for handheld devices is presented. The presented framework is enabled by using a server-client data communication structure, in which the search for tracking targets among a database of images is performed on the server-side while pixel-wise 3D tracking is performed on the client-side, which, in this case, is a handheld mobile device. Image search on the server-side adopts a residual-enhanced image descriptors representation that gives the framework a scalability property. The tracking algorithm on the client-side is based on a gravity-aligned feature descriptor which takes the advantage of a sensor-equipped mobile device and an optimized intensity-based image alignment approach that ensures the accuracy of 3D tracking. Automatic content streaming is achieved by using a key-frame selection algorithm, client working phase monitoring and standardized rules for content communication between the server and client. The recognition accuracy test performed on a standard dataset shows that the method adopted in the presented framework outperforms the Bag-of-Words (BoW) method that has been used in some of the previous systems. Experimental test conducted on a set of video sequences indicated the real-time performance of the tracking system with a frame rate at 15-30 frames per second. The presented framework is exposed to be functional in practical situations with a demonstration application on a campus walk-around.
Abstract: This paper discusses the technical challenges in
maritime image processing and machine vision problems for video
streams generated by cameras. Even well documented problems
of horizon detection and registration of frames in a video are
very challenging in maritime scenarios. More advanced problems
of background subtraction and object detection in video streams
are very challenging. Challenges arising from the dynamic nature
of the background, unavailability of static cues, presence of small
objects at distant backgrounds, illumination effects, all contribute to
the challenges as discussed here.
Abstract: Utilization of solar energy can be found in various domestic and industrial applications. The performance of any solar collector is largely affected by various parameters such as glazing, absorber plate, top covers, and heating pipes. Technology improvements have brought us another method for conversion of solar energy to direct electricity using solar photovoltaic system. Utilization and extraction of solar energy is the biggest problem in these conversion methods. This paper aims to overcome these problems and take the advantages of available energy from solar by maximizing the utilization through solar tracking system using a refrigerant as a working medium. The use of this tracking system can help increase the efficiency of conversion devices by maximum utilization of solar energy. The dual axis tracking system gives maximum energy output compared to single axis tracking system.
Abstract: In this paper, there is an implementation, verification, and graphical demonstration of a software application, which can be used swiftly over different preliminary orbit determination methods. A passive orbit determination method is used in this study to determine the location of a satellite or a flying body. It is named a passive orbit determination because it depends on observation without the use of any aids (radio and laser) installed on satellite. In order to understand how these methods work and how their output is accurate when compared with available verification data, the built models help in knowing the different inputs used with each method. Output from the different orbit determination methods (Gibbs, Lambert, and Gauss) will be compared with each other and verified by the data obtained from Satellite Tool Kit (STK) application. A modified model including all of the orbit determination methods using the same input will be introduced to investigate different models output (orbital parameters) for the same input (azimuth, elevation, and time). Simulation software is implemented using MATLAB. A Graphical User Interface (GUI) application named OrDet is produced using the GUI of MATLAB. It includes all the available used inputs and it outputs the current Classical Orbital Elements (COE) of satellite under observation. Produced COE are then used to propagate for a complete revolution and plotted on a 3-D view. Modified model which uses an adapter to allow same input parameters, passes these parameters to the preliminary orbit determination methods under study. Result from all orbit determination methods yield exactly the same COE output, which shows the equality of concept in determination of satellite’s location, but with different numerical methods.