YorkSpace has migrated to a new version of its software. Access our Help Resources to learn how to use the refreshed site. Contact diginit@yorku.ca if you have any questions about the migration.
 

Earth & Space Science

Permanent URI for this collection

Browse

Recent Submissions

Now showing 1 - 20 of 168
  • ItemOpen Access
    Image-Based Spatial Change Detection Using Deep Learning
    (2023-12-08) Bousias- Alexakis, Evangelos; Armenakis, Costas
    Image change detection is an invaluable tool in monitoring and understanding the built and physical environments and supporting decision-making. Many recent research approaches for automatic change detection have been based on Deep Learning (DL) techniques and especially on variations of Convolutional Neural Network (CNN) architectures. CNNs have achieved notable success thanks to their great representational capacity, straightforward training, and state-of-the-art performance in visual tasks. Nevertheless, CNNs, like most DL approaches, still face limitations relating to their reliance on extensive labelled datasets, the localization accuracy and detail of the predicted outputs, and the notable model performance degradation when the target data have different characteristics from the training data. This research contributes to the development and evaluation of novel DL methods and algorithms for automated image-based spatial change detection. It investigates novel architectures for enhanced model performance and ways to address the limited availability of labelled data and improve model generalizability. Two main approaches are investigated: (i) a direct approach, where both instances are fed into the DL model that is trained to output the prediction of the change map, and (ii) a post-classification approach, where change detection is performed based on each epoch’s semantic segmentation predictions. In both cases, novel, enhanced CNN architectures have been proposed that leverage the semantic information of objects’ boundaries to improve the accuracy of the model’s predictions. Furthermore, training frameworks inspired by self-ensembling and the Mean Teacher method were developed for semi-supervised learning and domain adaptation, attenuating the models’ reliance on large, labelled training datasets and improving their generalization performance. We evaluated the proposed approaches on multiple datasets using the precision, recall, F1 score, and Intersection over Union (IoU) metrics. The results indicate that both boundary-enhanced approaches lead to consistent, albeit marginal, benefits between 1 and 2 %. Notably, the proposed semi-supervised training framework for direct CD approximately matches the performance of the fully supervised approach while using only 20% of the available training labels. In domain adaptation, our post-classification approach significantly outperforms typical supervised methods, with the most notable gains in recall rate (>22%) and IoU (12.6%). These findings highlight the effectiveness of our techniques.
  • ItemOpen Access
    An Examination of the ThermopiezoelectricEffect in Piezoelectric Actuators Via Finite Elements
    (2023-12-08) Toledo, Rafael Chaves; Orszulik, Ryan
    The application of piezoelectric actuators in smart structures is a rapidly developing field, especially in aerospace environments. Since thermal effects play an important role in aerospace environments, thermopiezoelectricity has been studied as it takes into account the thermal field in addition to the mechanical and electrical fields. As a result, the coupling effects among these three fields have to be considered, including the pyroelectric (change in the electric potential due to the presence of a thermal field) and electrocaloric (change in the temperature when an electric field is applied) effects. This thesis presents an examination of how these coupled effects can affect the performance of piezoelectric bender and stack actuators in varying external environments. More specifically, this thesis investigates the influence of the pyroelectric and electrocaloric effects on the positioning and dynamic performance of these actuators in static and dynamic cases by using a custom written finite element code that considers the three fully coupled field equations of thermopiezoelectricity.
  • ItemOpen Access
    Visual-LiDAR Simultaneous Localization and Mapping
    (2023-12-08) Ahmadi, SeyedMostafa; Sohn, Gunho
    Simultaneous Localization And Mapping (SLAM) has garnered significant attention in robotics research over the years. While SLAM has demonstrated success, its application in mobile mapping systems (MMS) presents unique challenges. This study builds upon prior research (RPV-SLAM), extending its framework to enhance accuracy and perform boundary tests on a specific MMS, Maverick MMS. Our contribution introduces a novel SLAM approach termed HDPV-SLAM, addressing critical limitations encountered by the existing system. The first challenge addressed is the sparsity of LiDAR depth data, complicating its correlation with extracted visual features from RGB images. The second challenge stems from the lack of horizontal overlap between the panoramic camera and the tilted LiDAR sensor, causing difficulties in depth association. Furthermore, a comprehensive dataset named YUTO MMS is presented to the public. This dataset spans 18.95 km and was collected from diverse environments, including York University's Keel campus and Teledyne Optech headquarters building.
  • ItemOpen Access
    Agent-Based Modelling and Simulation of Sidewalk Delivery Robots' Interaction with Pedestrians
    (2023-12-08) Hassan, Ali; Sohn, Gunho
    In the evolving urban landscape, the surge of Sidewalk Autonomous Delivery Robots (SADRs) calls for insights into their pedestrian interactions. This thesis explores these dynamics and presents a novel web-based simulator, "TwinWalk", to emulate such interactions. Rooted in a GIS-enhanced digital twin of a campus-like urban setting, TwinWalk employs agent-based modeling, steering behaviours, and the Predictive Avoidance Model (PAM) to illustrate collision scenarios and human-robot interactions. Experiments reveal that while SADRs maintain safety buffers, they pose collision risks. Notably, pedestrians often breach safety distances, implicating them in proximity issues. This emphasizes the need for further research and specialized safety measures. The simulator aids urban planners and researchers in assessing design interventions and policies. The study suggests that enlarged safety zones around robots can reduce collisions and enhance pedestrian flow, promoting harmonious robot-human urban coexistence.
  • ItemOpen Access
    A Study of Buried Organics on Mars - A Computational And Experimental Approach
    (2023-12-08) Das, Ankita; Sapers, Haley; Moore, John
    This thesis consists of two related projects investigating exogenous organic burial on the surface of Mars and potential detection methods. Mars receives a significant quantity of organic material through exogenous sources such as micrometeorites. This work investigates the preservation of such material within sedimentary structures formed through the gradual airfall of dust particles. The computer model devised indicates rapid burial of organics and presents amount of organic carbon preserved at selected locations on Mars after 10 Martian years (99% preserved). The model is further extended globally to point at sites on Mars which are most favorable for organic preservation and hence possible sites for robotic exploration. The experimental section of this thesis investigates organic detection (Tryptophan) in simulated Martian dust matrix (JSC-1 Mars) through Ultraviolet Light Induced Fluorescence (LIF) spectroscopy. Instrument used was unable to detect Tryptophan in 100 ppm concentrations owing to the UV absorptive nature of JSC-1 Mars.
  • ItemOpen Access
    A Novel Controlled Environment Study of Prebiotic Organic Material in the Tagish Lake Meteorite Using Raman and Fluorescence Spectroscopy: Implications for Asteroid-Return Sample Analysis
    (2023-12-08) Lymer, Elizabeth Anne; Daly, Michael
    Several ambitious missions such as OSIRIS-REx, Hayabusa2 and Mars Sample Return are currently collecting or planning to collect the most pristine planetary samples to analyze on Earth since the Apollo missions. Therefore, it is essential to create instrumentation that will allow preliminary compositional analysis of these materials without the need for sample preparation and in a minimally destructive way, while preventing oxidation, alteration, or contamination from Earth’s environment. At York University, a small, transportable environmental chamber with temperature, pressure, and atmosphere control has been integrated on an optical table with a combined UV (266 nm) and Green (532 nm) Raman, laser-induced fluorescence, and time-resolved laser-induced fluorescence instrument. This system can collect high-resolution 2D spectroscopic maps, long accumulation point analysis and time-resolved fluorescence ‘fingerprinting’ of minerals and organics. The sample chamber is capable of pressures < 10E-4 mbar, temperatures < -20C, and different atmosphere compositions such nitrogen, argon or carbon dioxide via gas hookup. The intended use of the system is to detect and identify minerals and organics within sensitive or fragile planetary samples using minimally destructive laser-based techniques, while maintaining specific environmental conditions to preserve and maintain the pristineness of the material being analyzed, such as samples returned from planetary missions. To validate the functionality of the system as a laboratory tool for samples returned from space or for planetary science research, several experiments were conducted using different materials to showcase the different modalities of the instrumentation. One such experiment is conducted on a non-pristine piece of the Tagish Lake meteorite, containing some of the most primitive materials in our solar system. The meteorite fragment was kept frozen and in an inert atmosphere during Raman (UV and green) and Fluorescence (UV) analyses. These spectroscopic techniques resulted in high-resolution 2D maps of the surface of the meteorite fragment, showing mineral localities and groupings, as well as organic constituents present. UV-Raman in particular results in spectra containing aromatic and aliphatic hydrocarbon, ketones, and cyano radical peaks, all of which are the basic building blocks of important organic constituents like amino acids which play an integral role in biotic life on Earth. The detection of such compounds in the Tagish Lake meteorite using UV-Raman spectroscopy is novel and provides a unique tool to analyze organic constituents within sensitive materials in a very minimally destructive way with no sample preparation.
  • ItemOpen Access
    Advanced Data Fusion Methods to Improve Wetland Classification Using Multi-Source Remotely Sensed Data
    (2023-12-08) Judah, Aaron Jonathan; Hu, Baoxin
    The goal of this research was to improve wetland classification accuracy and the reduction of classification errors and uncertainty by fully exploiting multi-source remotely sensed, and ancillary data and image metrics using advanced data analysis techniques. This PhD research executed in three phases: 1. Explorations of data type selections and significance in support of wetland classification. 2. The development of a hierarchically-based classification approach to best exploit the data identified and characterized through the first study. 3. The development of an ensemble classifier incorporating the aforementioned developments with Dempster-Shafer (D-S) theory in order to reduce errors and streamline computations. The first phase explored the most effective data features, and metrics or families of data features in support of wetland classification. It was found that wetlands were best classified using the NDVI calculated from optical imagery obtained in the spring months, radar backscatter coefficients, surface temperature, and ancillary data such as surface slope, computed through either a Random Forest (RF) or Support Vector Machine (SVM) classifier. This work was also able to produce a wetland land cover map with an accuracy of 87.51% - an improvement from the ~82% typical of similar datasets and landcover types. In the second phase a more effective approach to classify the aforementioned features in order to fully utilize the discriminant power of those features was explored. This was done through two hierarchically based classification strategies. The second hierarchically based RF classification methodology produced the most accurate classification result (91.94%). The third phase focused on how to better exploit broad class separations and to reduce the propagation of errors and uncertainty which cascades through the classification hierarchy. These classifiers were integrated using D–S theory. Classification resulted in an overall accuracy of ~93% an improvement of 5% when compared to a traditional classification method. High level of confidence (>85%) misclassified pixels were reduced by ~10%. The major contribution of this research was the improvement of classification accuracy and the reduction of classification errors and uncertainty through use of multiple classifiers, designed to best exploit broad class separations, through selected data features computed within a D-S framework.
  • ItemOpen Access
    Quantifying Methane Fluxes on Mars: Modeling the Vertical Evolution of Martian Methane for Improved Detection and Analysis
    (2023-12-08) Walters, Madeline Elise; Moores, John; Gordon, Mark
    Searching for the presence of biosignatures on Mars is central to understanding and discovering clues to the geological and chemical history of the planet, and may lead us to unexpected discoveries. Methane has been observed in the Martian atmosphere for several decades, however, our understanding of the behavior of methane on Mars remains limited, particularly in terms of its diurnal variations, and its sources and sinks. On Earth, methane is regulated largely by biological activity, thus the presence of methane on Mars raises the question of whether the source of Martian methane is geochemical or biogenic in origin. Measurements of methane by the Sample Analysis at Mars Tunable Laser Spectrometer (SAM-TLS) aboard NASA's Curiosity Rover have revealed a background seasonal cycle of methane, though no measurement of the full diurnal cycle have yet been made. We use global circulation model (GCM) outputs to model the diurnal vertical evolution of methane on Mars and the strength of its flux from the surface. We also look at the allowable flight temperatures (AFT) for a dedicated instrument that is able to sample the Martian atmosphere hourly at sub-ppbv levels to better characterize the Martian methane cycle. The precision of such an instrument relies on the cooling of the gas sample, which must be kept free of condensing water vapor. Therefore, we look at the variability of the water frost point which reveals the areas of the Martian surface at which the instrument is most sensitive.
  • ItemOpen Access
    Influence of Horizontal Model Resolution on the Spatial Scale of Extreme Precipitation Events
    (2023-12-08) Ali, Syed Muhammad Anas; Tandon, Neil
    Previous work has shown that strong ascending motion is a key driver of extreme precipitation events (EPEs). Thus, the horizontal spatial scales of this “extreme ascent” are likely important for determining the spatial scales of EPEs. Therefore, understanding how climate models capture horizontal scales of ascending anomalies is critical to understanding and assessing climate models’ simulations and projections of extreme precipitation. Analyzing daily output from 27 models participating in the Coupled Model Intercomparison Project phase 6 (CMIP6) and High-Resolution Model Intercomparison Project (HighResMIP), we show that horizontal model resolution is a key influence on the horizontal scales of extreme ascent. We compute the horizontal scale for a given EPE as the e-folding distance of the vertical velocity anomaly on the day of the EPE, which is scaled to produce an inverse wavenumber. We then composite these horizontal scales over all annual maximum EPEs between 1981 and 2000 for each model. We focused on the horizontal scale zonally averaged over the 40S-55S latitude band. Our analysis suggests that model horizontal resolution places an upper limit on the horizontal scale of extreme ascent. Models with around 150 longitude points (approximately 153 km resolution at 55S) have mean horizontal scales topping out at approximately 320 km, and this upper limit decreases to approximately 220 km for models with 500 longitude points (approximately 46 km resolution at 55S). Additional analysis shows that the horizontal scales for geopotential anomalies during EPEs have no clear resolution dependence. However, the horizontal scales of geopotential were generally larger (700-1100 km) than those for vertical velocity or precipitation anomalies, and more in line with theoretical expectations based on the Rossby radius of deformation. Additional insight is gained through analysis of grid-scale and convective precipitation. Altogether, these results suggest that the simulated large-scale dynamics associated with EPEs is realistic, but the models are convecting at the grid scale rather than sub-grid scale. This is unrealistic as convective precipitation is expected to contribute strongly to extreme precipitation events. Additional analysis ensured that our results were not just a product of grid box storms, and that they were not sensitive to internal variability or temporal resolution.
  • ItemOpen Access
    CMOS Imager for Nanosatellite Applications
    (2014-10) Irvin, Patrick Edward Garret; Lee, Regina S. K.
    This research examines the capabilities of Commercial-off-the-shelf CMOS image sensors for use on nanosatellites as a star tracker. An emphasis is placed on selecting low cost components that meet the restrictions on mass and power of nanosatellites. To determine the expected sky coverage a Monte Carlo simulation is used with different limiting magnitudes and orbits. An image sensor, and lens system are selected and are interfaced using a Field Programmable Gate Array for data processing. The camera is then characterized in lab before being tested imaging the night sky.
  • ItemOpen Access
    Case Study of Foehn Events Over Alborz Mountains in Iran
    (2022-08) Sepehri, Jafar; Taylor, Peter
    Warm and dry foehn winds can significantly affect human life in mountainous regions worldwide. Foehn’s features include rising temperature, dropping relative humidity, and a fixed direction high wind commencement. Iran, a mountainous country, is affected by foehn in many places. The northern area of Iran is bordered by the Caspian Sea in the north and the Alborz Mountains in the south. The southwest part of the Caspian Sea coastal area is studied here. Large-amplitude mountain waves have occurred and caused severe wildfires in the Caspian forests. This study demonstrates that foehn events can occur due to high pressure over the interior regions and a lee cyclone over the southern Caspian Sea, with a strong south-north pressure gradient across the Alborz Mountains. Four typical 2021 wind events will be used to illustrate this foehn situation. Sample simulation results indicate the performance of WRF in predicting the variations of meteorological quantities.
  • ItemOpen Access
    Efficient GNSS Signal Acquisition Method for GNSS/GNSS-R Software-Defined Receivers
    (2023-08-04) Samigulin, Shamil Nailevich; Bisnath, Sunil; Lee, Regina
    This thesis presents a novel acquisition algorithm for Global Navigation Satellite Systems (GNSS) that can be efficiently implemented on small digital devices such as software defined radios (SDRs) field programmable gate arrays (FPGAs). The algorithm is designed to improve the performance of GNSS signal acquisition for applications in GNSS reflectometry (GNSS-R), a remote sensing technique that uses GNSS signals as a source of information. Using pre-acquisition processing and partial correlation, the proposed algorithm reduces the computational complexity of conventional GNSS acquisition methods by ~ 27 times, making it suitable for such low-cost devices. The thesis begins by introducing the GNSS technology and its spectrum, followed by a review of existing acquisition algorithms and their application in GNSS reflectometry. The novel acquisition strategy is developed, and its performance discussed, along with opportunities for future work.
  • ItemOpen Access
    FPGA-based GNSS receiver design for reflectometry applications
    (2023-08-04) Guruprasad, Surabhi; Bisnath, Sunil; Lee, Regina
    Research has shown that Global Navigation Satellite System (GNSS) signals reflecting off the Earth’s surface can be detected by receivers in low Earth orbit (LEO). The weak reflected signal properties are analyzed to characterize geophysical properties such as soil moisture, sea surface height, ocean surface wind speed and sea ice detection. This method of remote sensing is known as GNSS reflectometry (GNSS-R). Commercial GNSS receivers have historically been designed to only detect and process direct GNSS signals and cannot be repurposed to collect relevant data for reflectometry. Data collected by some orbital receivers have been made public; however, due to the high volume of data, data are truncated and are insufficient for science applications. Therefore, to be able to develop and test new algorithms, a custom GNSS-R receiver is designed and implemented. The developed GNSS-R receiver is implemented using field programmable gate array (FPGA) technology. The GNSS-R receiver prototype uses 1-bit signal resolution resulting in a compact design requiring minimal FPGA resources. Several field results show that the receiver prototype can successfully track direct and reflected GNSS signals in real-time. The observations indicate that the carrier-to-noise density ratio of signals reflecting from the surface of water was on average approximately 7 dB higher than the C/N_0 recorded when tracking land reflections. The difference in C/N_0 between water and land reflections is significant enough to conclude that a GNSS-R receiver using a one-bit quantization GNSS signal can be used for reflectometry applications. To increase the sensitivity of the receiver to weak reflected signals (-140 dBm), the FPGA-based signal processing module is enhanced using the alternate half-bit method. The receiver sensitivity improved from -35 dB to -46 dB (signal-to-noise ratio). Another challenge for GNSS-R receivers is generation of high-resolution delay Doppler maps (DDMs). DDMs provide insight into the reflecting surface characteristics based on signal scattering. The developed GNSS-R receiver demonstrates that high resolution DDMs can indeed be generated in real-time using 1-bit GNSS signal resolution. Data collected by the CYGNSS spacecraft are used to test and validate the receiver implementation. Results show that distinct DDMs are generated for land and water reflections and correlates (in terms of SNR) with research conducted using 2-bit signal resolution.
  • ItemOpen Access
    Examining the Atmospheric Transport of Microplastics using the HYSPLIT Model
    (2023-08-04) Ward, Eric Bradley; Gordon, Mark; Hanson, Ronald
    Microplastics are tiny particles less than 5 mm in size and are a growing scientific concern, given the potential harm caused across ecosystems as plastic use increases globally. To further understand the atmosphere's role, the HYSPLIT model was utilized to identify differences in transport distance and deposition area by particle size and shape. The extent of microplastic transport and deposition varied significantly by shape for particles larger than 6 µm. Long fibres deposited over a 32% greater area than spherical particles at the largest size of 23.5 µm. The maximum deposition area occurred at 4.5 µm, varying in area by less than 0.25% by shape. As particles smaller than 10 µm have the largest potential to cause adverse health effects, accurately modelling the shape of atmospheric microplastic transport is crucial to determining the range and amount of deposition globally, especially in the 6 µm to 10 µm size range.
  • ItemOpen Access
    Space Traffic Camera as an Opportunistic Sensor Towards Real-Time Space Domain Awareness
    (2023-08-04) Dave, Siddharth Ghanshyam; Lee, Regina
    The space domain is a congested and contested environment that requires constant surveillance for risk assessment. The adequacy of current and future space domain awareness is limited by our ability to collect quantitative information of the events in low Earth orbit. A network of scalable and geographically distributed sensors has the potential to collect a large volume of data for a more complete sky coverage than is possible today. Classic terrestrial and space surveillance sensors are too large and unsuitable to scale for better distribution and coverage. Spacecraft mounted star trackers present a novel opportunity to collect valuable surveillance data using commercially-proven technologies. Research presented in this dissertation evaluates scalability and detection performance, establishes an image processing and data compression framework, and describes the experimental validation of a prototype sub-orbital mission. The dissertation concludes that the potential data collection capacity of space traffic cameras, adopted from star trackers, are feasible to achieve the desired data volume for improved low Earth orbit surveillance. The sensor network discussed can provide novel access with improved re-visit times to detect, track and classify resident space objects to augment existing \gls{sda} capabilities. This conclusion is presented in the context of data collection utility and quantity, data processing and means of centralization, and sensor design and feasibility survey. The conclusion is further supported with the knowledge that the necessary hardware required for data collection, i.e., the current count of active star trackers in Earth orbit, is already in place, and is expected to multiply due to increasing launch cadence. The results also suggest that such a network of space traffic cameras is a more cost effective approach for space surveillance, as the design presented can be automated and scalable to large constellations. The data collected can be used for space object characterization studies for development of in-orbit services in addition to spacecraft custody operations. The research concludes with recommendations for future work that further advance the usage of existing low-cost commercial hardware for space domain awareness. Contributions from this research demonstrate a sustainable method to achieve commercial space safety and remote services in low Earth orbit.
  • ItemOpen Access
    Continuous Urban Navigation with Next-Generation, Mass Market Navigation Sensors and Adaptive Filtering
    (2023-08-04) Vana, Sudha; Bisnath, Sunil B.
    The Global Navigation Satellite System (GNSS) Precise Point positioning (PPP) technique benefits from not needing local ground infrastructure such as reference stations and accuracy attained is at the decimetre-level, which approaches real-time-kinematic (RTK) performance. However, due to its long position solution initialization period and dependence on the receiver measurements, PPP finds limited utility in obstructed areas. The emergence of low-cost, high-performance micro-electromechanical sensor (MEMS) inertial measurement units (IMUs) has prompted research in integrated navigation solutions with GNSS PPP augmentation. In this study, novel research is performed using a low-cost, dual- and triple-frequency (DF and TF) GNSS and, MEMS IMU to attain decimetre to sub-metre accuracy in challenging environments. New-generation applications demand decimetre-level positional accuracy while using low-cost equipment. PPP that does not need any local infrastructure has become a promising technique to be used for such mass-market applications. The objectives of the research are to examine the effect of sensor constraining to improve position accuracy, assess the performance of TF PPP and MEMS IMU algorithm in open-sky and simulated outages, and use adaptive filtering to maintain decimetre to sub-metre-level accuracy in all environments using low-cost sensors. An uncombined GNSS PPP solution was integrated with MEMS IMU using tightly-coupled architecture. The novelty addressed by this research is the combination of the low-cost hardware and the software constraining that is used together to provide significantly improved continuous and accurate navigation in the urban environment, which has not been examined previously in the PPP + IMU research area. Furthermore, to achieve sub-metre level horizontal accuracy, modification to the traditional robust adaptive Kalman filter (RAKF) is proposed. Data collected in open sky, as well as urban environments, were assessed for the performance in simulated as well as real urban outages. The integrated system performs with less than a decimetre-level accuracy in open-sky and sub-metre-level in simulated environment. Sub-metre-level rms results were attained by using the novel modified RAKF in urban areas. The outcomes of this research are reassuring towards achieving continuous navigation solutions with decimetre to sub-metre level accuracy for modern applications that demand higher accuracy in all environments while using low-cost equipment.
  • ItemOpen Access
    Global, Instantaneous, Centimetre, Satellite-Based Positioning with Precise Point Positioning
    (2023-08-04) Naciri, Nacer; Bisnath, Sunil B.
    Real-Time Kinematic (RTK) has been the reference technique when it comes to precise (centimetre-level) positioning for Global Navigation Satellite Systems (GNSSs). The fact that RTK performance depends heavily on the distance between the receiver and the base station has led to the technique being deployed only regionally, in areas where GNSS reference station infrastructure is present. Precise Point Positioning (PPP) is a global solution by design, though requiring tens of minutes to reach RTK levels of performance. In this research, attempts are made to improve PPP performance to reach RTK-type performance. Given that PPP is a measurement-dependent technique, this research starts with making use of signals from all four major GNSS constellations: GPS, GLONASS, Galileo, and BeiDou-2/3. These constellations are coupled with the use of up to four frequencies, as opposed to only two frequencies as is typical for PPP. Carrier-phase ambiguities are resolved to their integer values on all processed signals as well. A model is derived from first principles to process all these measurements and fix their ambiguities. Results show that incorporating four frequencies has great benefits in improving user positions and increasing the likelihood of correct fixing of ambiguities due to the correlation that exists between ambiguities from the same satellite. Results demonstrate that PPP can reach regional RTK levels of performance with only using global corrections, as instantaneous convergence to 2.5 cm error is achieved consistently, making of PPP a truly global precise positioning technique. With these results in mind, corrections from Galileo’s High Accuracy Service (HAS) are analyzed, where, for the first time on a global scale, corrections are being transmitted by a GNSS constellation. Thanks to early access to these signals, an in-depth analysis of the corrections and user performance is carried out. Good performance is found to be achieved with these limited test signals, showing that PPP can become the de facto global precise positioning technique. In cases where PPP needs to be augmented, a proof-of-concept is proposed where a global PPP solution is augmented with ionospheric corrections generated from an NRTK network. The generated corrections are found to improve PPP solution by reducing convergence time.
  • ItemOpen Access
    Deep Convolutional Neural Network Based Single Tree Detection Using Volumetric Module From Airborne Lidar Data
    (2023-03-28) Lee, Hyungju; Sohn, Gunho; Ko, Connie
    There was an undeniable success of Deep Learning networks for visual data analytics such as object detection and segmentation in recent years, while the adaptation to tree detection has been rare. In this paper, we pursue to achieve individual tree identification, defined as a detection of an individual tree as each object, with deep convolutional neural networks to create and update tree inventories using LiDAR information. The first objective was to provide a suitable dataset that can be used to test such networks and to create a module that attempts to increase the 3D object detection algorithms' detection accuracy. This novel dataset was created by fusing LiDAR data gathered by Teledyne Optech with field data collected by York University. The second was to develop an appropriate accuracy increasing volumetric module. For this module, the learnable weights concept was introduced, which enable to increase detection precision of the object detection algorithm.
  • ItemOpen Access
    Development of Advanced Remote Sensing Methods in Quantifying Wildlife Habitat Management
    (2023-03-28) Zhang, Wen; Hu, Baoxin; Brown, Glen
    Wildlife habitats have been affected by human activities and climate change. Animal diversity is declining at an unprecedented rate. Tools used to obtain a rapid assessment of wildlife habitats at different scales are urgently needed. The habitat management tools that are currently used for conservation and monitoring wildlife are often limited by the availability of mapped habitat information that is tailored to the wildlife of interest and that covers appropriate geographic and temporal extents of interest. Failure to adequately map specific habitat features can limit effective management. Advancements in remote sensing and related technologies have increased the resolution and quantity of landscape data, providing an excellent opportunity to extract various environmental features for examining habitat selection and mapping wildlife habitats to a broad extent. To exploit the potential of the emergent remote sensing data sets, the focus of this study was to develop advanced methodologies to derive information related to the properties of environmental features at different scales and to generate tools to improve the understanding of a wildlife habitat landscape that can benefit from habitat management. Specifically, an advanced algorithm was developed that utilized spatial pattern analysis to classify the forest succession stages from optical imagery and had a classification accuracy of 89%. In addition, a novel method was proposed to extract road features from the road structure knowledge followed by a deep learning VGG 16 classification for a refined output. An overall accuracy of 74% was achieved for the forest road extraction. A robust and operational stepwise automatic thresholding method was developed to accurately map the dynamics of surface water bodies from SAR data, with an overall accuracy of 95%. In addition, an advanced fuzzy AHP model was utilized to accurately map beaver-altered wetlands in the landscape using remote sensing products derived based on the knowledge of beaver activities, where an average of 83.0% of the known beaver dams and 72.5% of the known beaver ponds were correctly identified. In conclusion, this research demonstrated that the advanced methods utilizing multi-source and multi-temporal remote sensing data could effectively characterize and extract environmental features that benefit wildlife habitat management.
  • ItemOpen Access
    Active Reinforcement Learning for the Semantic Segmentation of Images Captured by Mobile Sensors
    (2023-03-28) Jodeiri Rad, Mahya; Armenakis, Costas
    Neural Networks have been employed to attain acceptable performance on semantic segmentation. To perform well, many supervised learning algorithms require a large amount of annotated data. Furthermore, real-world datasets are frequently severely unbalanced, resulting in poor detection of underrepresented classes. The annotation task requires time-consuming human labor. This thesis investigates the use of a reinforced active learning as region selection method to reduce human labor while achieving competitive results. A Deep Query Network (DQN) is utilized to identify the best strategy to label the most informative regions of the image. A Mean Intersection over Union (MIoU) training performance equivalent to 98% of the fully supervised segmentation network was achieved with labeling only 8% of dataset. Another 8% of labelled dataset was used for training the DQN. The performance of all three segmentation networks trained with regions selected by Frequency Weighted Average (FWA) IoU is better in comparison with baseline methods.