Signal Processing Commons^™

Estimation Of Atmospheric Conditions Over A Long Horizontal Path Using Multi-Frame Blind Deconvolution (Mfbd) Techniques In Comparison With Delayed Tilt Anisoplanatism (Delta) Software, Hannah Stoll

Dissertations, Master's Theses and Master's Reports

The potential to track and view objects in space from the ground with greater near real time knowledge of the intervening turbulence would be a revolutionary capability. The objective of this thesis is to cross-validate two separate methods used to estimate the Fried parameter. This verification is a step toward a commercial grade product that would make real-time estimates of the turbulence strength along an optical path from a ground-based observatory to a satellite in orbit around the Earth. Michigan Technological University has developed a multi-frame blind deconvolution (MFBD) algorithm used to estimate r0 and it was tested against MZA’s …

Development Of A Software-Defined Underwater Acoustic Communication System, Zijian Zhu

Dissertations, Master's Theses and Master's Reports

This report started with a brief history and recent development of underwater acoustic communication systems as well as software-defined radio technologies. Then, some challenges from underwater acoustic channels and available underwater acoustic communication modems are discussed. After finished introducing the basics of SDR and GNU Radio, a detailed description of implementing a software-defined acoustic communication system in GNU Radio are presented, along with some key concepts of the system.

Then, some hardware specifications are presented, following by detailed documentation on a software-defined acoustic communication system experiment with a host computer, a USRP, an acoustic hydrophone, and a hydrophone. At the …

Estimation Of Vehicle Trailer Rotation Using Calibrated Radars With Multiple Viewpoints, Kunle Olutomilayo

Dissertations, Master's Theses and Master's Reports

Multiple sensors are increasingly being deployed on systems for perception applications. In particular, vehicles are becoming equipped with a suite of sensors for advanced driver assistance features and autonomous driving. This dissertation considers automotive radar sensors mounted at the rear of a vehicle with the main objective of using their point cloud detections to estimate the rotation of a trailer which is attached to the vehicle's hitch ball. A simulation-based study of the problem is presented first. Thereafter, the problem is considered with respect to experimental radar data collected in both indoor and outdoor environments; the environmental difference is in …

Video And Image Super-Resolution Via Deep Learning With Attention Mechanism, Xuan Xu

Graduate Theses, Dissertations, and Problem Reports (ETD)

Image demosaicing, image super-resolution and video super-resolution are three important tasks in color imaging pipeline. Demosaicing deals with the recovery of missing color information and generation of full-resolution color images from so-called Color filter Array (CFA) such as Bayer pattern. Image super-resolution aims at increasing the spatial resolution and enhance important structures (e.g., edges and textures) in super-resolved images. Both spatial and temporal dependency are important to the task of video super-resolution, which has received increasingly more attention in recent years. Traditional solutions to these three low-level vision tasks lack generalization capability especially for real-world data. Recently, deep learning methods …

Kettlebell Ultra, Elissa Peters, Kathryn Wegman, Daniel Basch, Mason Pastorius

Williams Honors College, Honors Research Projects

This project will consist of an attachment to an average kettlebell that will track the number of repetitions that the user has performed. The device will send this data over Bluetooth to a smart phone application so the user can track their workout accurately.

Visual Music Assistant, David Klett

Williams Honors College, Honors Research Projects

The Visual Music Assistant system provides an augmented reality experience via Microsoft's HoloLens device. The application we will develop will provide an intuitive user interface to learn how to play on a keyboard (88-key piano).

Levitate, Stephen Hargreaves, Isaac Hammonds, Russel Pease, Joseph Morris

Williams Honors College, Honors Research Projects

Levitate is designed to create a healthier work environment by providing a way of entertaining a worker while not distracting them. To fulfill this task, the device needs to be interesting and self-sufficient while not being overly flashy. To complete this task, Levitate is designed so that it is able to lift a platform from rest using magnetism and that will be able to stabilize itself using electromagnets. It initiates this process by lifting the permanent magnet using servo motors. Then, the device is able to achieve stability through the use of hall sensors feeding their results to the board …

Digital, Automated Reactive Target System, Nicholas Haas, Saipranay Vellala, Trandon Ware, Thomas Martin

Williams Honors College, Honors Research Projects

In this era, technology is woven into almost every facet of our leisure activities. Although technology has innovated hobbies ranging from chess to soccer, the art of shooting has been neglected. Unnecessary insufficiency such as bullet ricochets off of mechanical steel targets, ineffective progress tracking, and general inaccessibility to outdoor training facilities are all improvable areas of this sport. The Dynamic Automated Reactive Target (D.A.R.T) System aims to fill some of these gaps and help modernize recreational marksmanship. Modeling the system after a dueling tree will optimize the use of the system and allow for different training models to challenge …

Bitbilliards, Grant Reinbolt

Williams Honors College, Honors Research Projects

The goal of the project is to create a billiards table capable of keeping track of the score while simultaneously racking the sunk balls under the table. To achieve this, image processing will be utilized to detect and determine which balls are above and below the table. Upon leaving the table, the balls will be held by gates to prevent the cue ball from entering the re-racking track. The balls will then enter the track and form a queue to be placed back into the triangle underneath the table. A gantry system consisting of an x and y axis will …

Aperture-Level Simultaneous Transmit And Receive (Star) With Digital Phased Arrays, Ian Cummings

Dissertations, Master's Theses and Master's Reports

In the signal processing community, it has long been assumed that transmitting and receiving useful signals at the same time in the same frequency band at the same physical location was impossible. A number of insights in antenna design, analog hardware, and digital signal processing have allowed researchers to achieve simultaneous transmit and receive (STAR) capability, sometimes also referred to as in-band full-duplex (IBFD). All STAR systems must mitigate the interference in the receive channel caused by the signals emitted by the system. This poses a significant challenge because of the immense disparity in the power of the transmitted and …

Articulatory-Wavenet: Deep Autoregressive Model For Acoustic-To-Articulatory Inversion, Narjes Alsadat Agha Seyed Mirza Bozorg

Theses and Dissertations--Electrical and Computer Engineering

Acoustic-to-Articulatory Inversion, the estimation of articulatory kinematics from speech, is an important problem which has received significant attention in recent years. Estimated articulatory movements from such models can be used for many applications, including speech synthesis, automatic speech recognition, and facial kinematics for talking-head animation devices. Knowledge about the position of the articulators can also be extremely useful in speech therapy systems and Computer-Aided Language Learning (CALL) and Computer-Aided Pronunciation Training (CAPT) systems for second language learners. Acoustic-to-Articulatory Inversion is a challenging problem due to the complexity of articulation patterns and significant inter-speaker differences. This is even more challenging when …

Calibration To Mitigate Near-Field Antennas Effects For A Mimo Radar Imaging System, Ha Hoang, Matthias John, Patrick Mcevoy, Max Ammann

Articles

A calibration method for a high-resolution hybrid MIMO turntable radar imaging system is presented. A line of small metal spheres is employed as a test pattern in the calibration process to measure the position shift caused by undesired antenna effects. The unwanted effects in the antenna near-field responses are analysed, modelled and significantly mitigated based on the symmetry and differences in the responses of the MIMO configuration.

Anomalous Event Detection And Localization Based On Deep Generative Adversarial Networks For Surveillance Videos, Thittaporn Ganokratanaa

Chulalongkorn University Theses and Dissertations (Chula ETD)

Anomaly detection is of great significance for intelligent surveillance videos. Current works typically struggle with object detection and localization problems due to crowded scenes and lack of sufficient prior information of the objects of interest during training, resulting in false-positive detection results. Thus, in this thesis, we propose two novel frameworks for video anomaly detection and localization. We first propose a Deep Spatiotemporal Translation Network (DSTN), a novel unsupervised anomaly detection and localization method based on Generative Adversarial Network (GAN) and Edge Wrapping (EW). In this work, we introduce (i) a novel fusion of background removal and real optical flow …

Design And Development Of Fiber Bragg Grating Sensor For Detecting Of Hydrogen Gas In Transformer Oil, Mohd Raffi Samsudin

Student Works (2020-2029)

Hermetically sealed oil-immersed transformers are the key to electricity distribution networks. They are also one of the most expensive facilities in the electricity supply network. Immediate replacement is expected for this type of transformer upon failure because it is directly connected to the customer. Transformer oil acts as insulation, coolant and condition indicator. To date, there is no economical real-time monitoring system available to monitor the transformer oil condition. Therefore, there is a need to develop a cheap, non-intrusive and non-electrical sensor to monitor the health of the transformer by measuring the amount of dissolved hydrogen gas in oil, which …

Ultra-Wideband Power Amplifier Design For Large Signal Application, Ragavan Krishnamoorthy

Student Works (2020-2029)

Broadband power amplifier design has become one of the most critical enabling block in today’s wireless communication technology. Numerous research efforts have been carried out throughout the years to establish high efficiency over wideband RF transmitter. This research work presents two approaches in achieving ultra-broadband power amplifier for RF transmitter. Firstly, a new technique for the design of ultra-broadband RF power amplifier is introduced, in which a combination of the reactance compensation and third-harmonic tuning are adopted. The design goal is to achieve 40 dBm (10W) output power across a wide frequency bandwidth operation. Theoretical design equations were developed to …

Comparative Study Of Model-Based And Learning-Based Disparity Map Fusion Methods, Douglas E. Kerr Jr.

Graduate Theses, Dissertations, and Problem Reports (ETD)

Creating an accurate depth map has several, valuable applications including augmented/virtual reality, autonomous navigation, indoor/outdoor mapping, object segmentation, and aerial topography. Current hardware solutions for precise 3D scanning are relatively expensive. To combat hardware costs, software alternatives based on stereoscopic images have previously been proposed. However, software solutions are less accurate than hardware solutions, such as laser scanning, and are subject to a variety of irregularities. Notably, disparity maps generated from stereo images typically fall short in cases of occlusion, near object boundaries, and on repetitive texture regions or texture-less regions. Several post-processing methods are examined in an effort to …

Object Identification In Radar Imaging Via The Reciprocity Gap Method, Matthew Charnley, Aihua W. Wood

Faculty Publications

In this paper, we present an experimental method for locating and identifying objects in radar imaging, specifically problems that could arise in physical situations. The data for the forward problem are generated using a discretization of the Lippmann‐Schwinger equation, and the inverse problem of object location is solved using the reciprocity gap approach to the linear sampling method. The main new development in this paper is an exploration of determining the permittivity of the object from the back‐scattered data, utilizing another discretization of the Lippmann‐Schwinger equation.
Abstract © AGU.

Image Restoration Using Automatic Damaged Regions Detection And Machine Learning-Based Inpainting Technique, Chloe Martin-King

Computational and Data Sciences (PhD) Dissertations

In this dissertation we propose two novel image restoration schemes. The first pertains to automatic detection of damaged regions in old photographs and digital images of cracked paintings. In cases when inpainting mask generation cannot be completely automatic, our detection algorithm facilitates precise mask creation, particularly useful for images containing damage that is tedious to annotate or difficult to geometrically define. The main contribution of this dissertation is the development and utilization of a new inpainting technique, region hiding, to repair a single image by training a convolutional neural network on various transformations of that image. Region hiding is also …

Visual Speech Recognition Using A 3d Convolutional Neural Network, Matthew Rochford

Master's Theses

Main stream automatic speech recognition (ASR) makes use of audio data to identify spoken words, however visual speech recognition (VSR) has recently been of increased interest to researchers. VSR is used when audio data is corrupted or missing entirely and also to further enhance the accuracy of audio-based ASR systems. In this research, we present both a framework for building 3D feature cubes of lip data from videos and a 3D convolutional neural network (CNN) architecture for performing classification on a dataset of 100 spoken words, recorded in an uncontrolled envi- ronment. Our 3D-CNN architecture achieves a testing accuracy of …

Radio Direction Finding Using Pseudo-Doppler For Uav-Based Animal Tracking, Anup Karki

Masters Theses

Radio Direction Finding (RDF) is commonly used for low cost tracking and navigation systems. However, for a low cost application and mobility, the design constraints are highly limited. Pseudo Doppler (PD) can improve RDF capabilities without being cost prohibitive. This work entails the analysis of PD RDF and its potential use for Unmanned Aerial Vehicles (UAV) that are currently employed in wildlife research animal tracking. PD is based on the doppler effect or doppler shift. The doppler effect works like a frequency modulator that increases or decreases the observed frequency depending on whether a signal source is approaching or receding …