Sharif Digital Repository

Human tracking is an essential step in many computer vision-based applications. As single view tracking may not be sufficiently robust and accurate, tracking based on multiple cameras has been widely considered in recent years. This thesis presents a distributed human tracking method in a smart camera network and introduces a particle filter design based on Histogram of Oriented Gradients (HOG) and color histogram. The proposed adaptive motion model also estimates the target speed from the history of its latest displacement and improves the robustness of the tracker by decreasing the probability of missing targets. In addition, a distributed data fusion method is proposed which fuses the... 

Measurement of human movement during work, sport as daily activities has important effects for rehabilitation and biomechanics experts.
Recently, several researches are concentrated on this technology significantly because of new progress in sensors and fusion sensors. In the near future, wearable and ambulatory devices will be used considerably in biomedical applications. In this study, we reviewed some methods about wearable measuring system then introduced and designed innovative wearable clothing for this purpose. Some specifications and advantages of our system in comparison with relevant researches are; using textile sensors that are manufactured through nano electro active polymer... 

The main subject of this project is to present a new algorithm for multiple RF sources localization by means of a group of robots. This problem (locating multiple radio sources) could be utilized in many domains, either military (passive radars) or civil (search and rescue) uses. Confronting detrimental effects of real environment on wave’s propagation pattern is the main challenge in this problem. Optimized movement of robots has been chosen as basic strategy to overcome this challenge. Presenting a general solution for this problem requires researches in four topics. The first one is simulation of environment and sensor with main objective of evaluating obtained algorithms’ performance.... 

A Synthetic Aperture Radar (SAR) is a system which collects the received signal from a ground target area along linear motion of a flying vehicle carrying it, in order to provide a very high resolution 2D image of it. In Inverse SAR (ISAR), the radar is fixed on the ground and the data collection is performed along the linear motion of the aerial target. Using a second receiver antenna close to the main transceiver antenna of ISAR, it is possible to make a three-dimensional (3D) image of the target, or equivalently find the 3D coordinates of the target scattering points. Such system is called bistatic, monopulse or interferometric ISAR (InISAR). In the conventional model of ISAR, the flying... 

The importance of the motion measurement systems is obvious due to survey the number of people who have joint pains through wrong movements. Considering the medical device’s trend toward wearable systems, the necessity of portable measurement instruments which are independent from experimental setup, is undeniable. Motion measurement systems based on MEMS Inertial sensors are widely used these days because they are cheap and portable. However these systems have some defects such as error accumulation and increasing error due to magnetic disturbance or high acceleration motions. Multisensory data fusion along with compensator filters is the best solution in order to reduce these errors.... 

Due to various applications of human motion capture techniques, developing low-cost methods that would be applicable in non-laboratory environments is under consideration. MEMS inertial sensors and Kinect are two low-cost devices that can be utilized in home-based motion capture systems, e.g. home-based rehabilitation. In this work, an unscented Kalman filter approach was developed based on the complementary properties of Kinect and the inertial sensors to fuse the orientation data of these two devices for human arm motion tracking during both stationary shoulder joint position and human body movement. A new measurement model of the fusion algorithm was obtained that can compensate for the... 

Real-time asynchronous data fusion for high-speed phenomena is an important and challenging task in the sensor networks. Examples of data fusion applications in sensor networks are: managing the traffic of maneuvering airplanes and ground vehicles in airside areas of an airport, traffic management in streets and roads, Driver Assistance Systems, guidance of antiaircraft and antimissile missiles. In all the data fusion applications the estimation of the required variables is necessary.
In this research two methods are introduced for real-time asynchronous data fusion, especially for track-to-track fusion of high-speed phenomena in sensor networks. The effectiveness and usability of these... 

Short-circuit fault detection is mandatory in a high voltage DC power supply to prevent fatal damage. The majority of converters employ a single sensor to detect the short-circuit fault. This attribute increases the interference vulnerability of the fault detection system in the presence of noise. Therefore, miss detections and false alarms are possible to occur. Miss detections and false alarms are harmful catastrophes in most applications. A commonly used method to suppress the noise impacts is using a low-bandwidth low-pass filter. However, the use of the low-bandwidth low-pass filter reduces the speed of fault detection due to the filter delay. This paper proposes a fast fault detection... 

Molecular communication (MC) is an emerging communication paradigm, whereas molecules are used as information carriers to establish communication among elements in nano-meter to meter scales. In this thesis, we investigate the problem of detecting and monitoring changes (abnormality) based on molecular communication, using quickest change point detection scheme. We assume the distributions and parameters of the system are known. To this end, we consider a network of multiple sensors, each sensing its surrounding and employing On-Off-keying modulation for data transmission toward a fusion center (FC). An abnormality initiates randomly in time and location, and further propagates in the... 

Estimating the remaining useful life (RUL) of critical assets is an essential task in system health management. It also has a crucial role in the optimization of maintenance scheduling and decision making about the future of the asset. On the other hand, since the rolling element bearings are widely used in the rotating machinery, estimation of their RUL is considered as remarkable progress in improving the reliability of the whole system. In this way, various models have been introduced to predict the RUL of rolling element bearings, which their proper functionality is affected by the underlying assumptions in the model structure. Because of the presence of several uncertainties and... 

The fire event may cause danger to life and property. In general, most fire-detection devices are set on the wall or ceiling. . However, this method is not flexible enough to detect a fire occurrence, and this needs the use of many fire-detection devices in a building. Since it is not suitable, we use an active fire-detection device to detect a fire incident. In the past, smoke sensors were used for fire detection, but often the smoke sensors triggered a false alarm when someone smoked in the room. Therefore, multisensor fire-detection methods are one of the current important developments for fire- detection technology. We design a fire-detection module using an ionization smoke sensor... 

This research focuses on the robust mobile robot localization exploiting motion information acquired from an optical mouse operating based on optical flow technology. Most techniques of visual motion measurement are based on the well research discipline called “optical flow”. Theoretically, optical flow as a method of localization can be highly accurate, but it is sensitive to the noise and surface texture/optical characteristics and distance variations between the CCD detector and surface. As one could not achieve acceptable results in practical situations, to handle these problems, we propose to attach an acceleration – gyro (INS) sensor on the CCD detector (optical mouse) to improve the... 

The Gaussian mixture (cardinalized) probability hypothesis density (GM-(C)PHD) filter is a closed form approximation of multi-target Bayes filter which can overcome most of multi-target tracking problems. Limited field of view, decreasing cost of cameras and its advances induce us to use large-scale camera networks. Increasing the size of camera networks make centralized networks practically inefficient. On the other hand, scalability, simplicity and low data transmission cost has made distributed networks a good replacement for centralized networks. However, data fusion in distributed network is sub-optimal due to unavailable cross-correlation.Among data fusion algorithms which deal with... 

Human motion capture is frequently used to study rehabilitation and clinical problems, as well as to provide realistic animation for the entertainment industry. IMU based systems as well as Marker based motion tracking systems are of most popular methods to track movement due to their low cost of implementation and lightweight. Results of IMU leads to unacceptable drift errors while marker based systems are Drift-free. Also, unlike cameras, IMUs are suitable for long-distance motion capturing. Based on the complementary properties of marker based systems and the inertial sensors, in this work, an Extended Kalman Filter approach was developed to fuse the data of two IMUs and a single camera... 

Given an event O and a set of experts E, we describe a method for finding a subset of experts S whose aggregated opinions best predict the outcome of O. Therefor, the problem can be regarded as team formation for performing a prediction task. In order to estimate competency of each team we propose measure Sum Squared Error which uses experts’ records of predictions during past k days. For simplicity, opinion pooling is selected as the method of information aggregation. we prove in case of simple averaging of opinions,finding best team is NP-hard. we suggest some rounding and heuristic algorithms for finding near optimal solutions. Simulation results show that a variation of tabu search used... 

In recent years health monitoring and prognostics of complex systems have been considered more than ever.. In the present researh, data - based approach has been selected among various prognostics and health monitoring approaches. One of the most challenging issues in data-based methods is how to map system sensor information to its health status. In this research, different methods of mapping are discussed. The results show that sensor fusion by principal component analysis (PCA) offers acceptable performance. This pattern produces a single-dimensional signal for health monitoring with high reliability.The second challenge is to predict the status and design of the prognostics module. For... 

Inertial motion capture systems widely use low-cost IMUs to obtain the orientation of human body segments, but these sensors alone are unable to estimate link positions. Therefore, this research used a SLAM method in conjunction with inertial data fusion to estimate link positions. SLAM is a method that tracks a target in a reconstructed map of the environment using a camera. This paper proposes quaternion-based extended and square-root unscented Kalman filters (EKF & SRUKF) algorithms for pose estimation. The Kalman filters use measurements based on SLAM position data, multi-link biomechanical constraints, and vertical referencing to correct errors. In addition to the sensor biases, the... 

Crop classification is one of the most important applications of remote sensing in agriculture. Knowing what crops are on the farm is invaluable both on a micro and macro scale. For example, this information can be used to design and imple- ment agricultural policies, product management and ensure food security. Also, this information can be used as a prerequisite for implementing other programs at the farm scale, such as monitoring and detecting anomalies during the crop growth cycle. Most of the studies in this field are focused on the optical data of the Sentinel-2 satel- lite, but the optical data are vulnerable to atmospheric conditions, and on the other hand, there is valuable... 

In image processing applications, sensors (Camera, LiDAR and Stereo) are essential for scene perception and Deep learning methods outperform most of the image processing tasks like 3D and 2D object detection and semantic segmentation. Different sensors are used in image processing tasks. Sensor fusion is using multiple sensors data to get better performance. Each sensor captures different data (e.g, color, texture, and depth). Some of them are distorted in inclement weather, intense illuminance changes, and dark environments which multi-sensor data fusion is used to overcome sensor weaknesses. One of the most important fields that sensor fusion used is Auto Driving cars (AD). Different...