BORDaR Bournemouth Online Research Data Repository
Not a member yet
216 research outputs found
Sort by
Holistic indoor scene understanding, modelling and reconstruction from single images
Get PDFOverall, this thesis provides a series of novel approaches towards holistic 3D indoor scene understanding, modelling and reconstruction. It aims at automatic 3D scene perception that enables machines to understand and predict 3D contents as human vision, which we hope could advance the boundaries of 3D vision in machine perception, robotics and Artificial Intelligence
Measuring spinal and trunk shape using an electromagnetic sensor
Get PDFA critical component in the clinical assessment of spinal and trunk disorders is the analysis of posture. Currently the gold-standard is restricted by repeated radiation exposure and whilst alternative surface methods are available, these are limited to detection of spinal shape only. To date, no surface method has been extended to also quantify trunk shape. In order to address this, the aims of this research were 1) develop a method for measuring spinal and trunk shape using an electromagnetic system; 2) determine the validity and reliability of this method and 3) explore the optimal data processing for this method.
Using a repeated measures design, data were collected on phantom models of different shapes using an electromagnetic system. This provided the three-dimensional co-ordinates from which spine and trunk angles were derived.
The 6th order polynomial fit was deemed optimal for spinal shape measurements with an electromagnetic system. These measurements were highly reliable (ICC = >0.999), highly repeatable (MDC = <0.018º, SEM = <0.007º) and shown to be valid compared to a flexicurve method. The Lowess function was recommended for trunk shape measurements as it yielded good-to-excellent repeatability (ICC = 0.809-0.999), high absolute reliability (MDC = 0.18-4.0º, SEM = 0.06-0.07º) and angles derived were valid compared to a flexicurve method.
This study addressed a clinical need by developing a novel method for measuring trunk shape in addition to spinal shape using a surface method which was shown to be valid and reliable. Exploration of the method’s optimal data processing techniques found the 6th order polynomial fit and Lowess function to be best for spinal shape and trunk shape measurements respectively. Additionally, whilst it is recommended that tangent lengths should not be used interchangeably, the tangent length chosen should not significantly affect measurements if used consistently. Meanwhile, the method’s non-invasive, non-ionising and low-cost features make it clinically attractive. Therefore, this research holds future prospects for the examination and monitoring of disease and treatment outcomes as well as, the understanding of many disorders, such as scoliosis. Although further research is warranted, this method has the potential for use in routine clinical practice
Generalised discriminative optimisation algorithms for augmented reality: theory and practice
No full textAugmented Reality (AR) technology achieves the seamless registration between virtual scenes and the real world. Three-dimensional registration is a core method for integrating virtual information and the real world. Most registration experiments are conducted on simple data models (such Standford Bunny model and the Bimba model) currently. Actually, augmented reality is significantly related to registration and object tracking in real scenes. In order to verify the better performance of the proposed algorithms, we created the point clouds of real scenes to evaluate the performance of the proposed algorithms on object tracking and the registration of scenes and models. We use KinectV2 to attain the depth images and RGB images of the moving models (bunny model, chicken model and parasaurolophus model ) standing in the hand of a people. Then point clouds are reconstructed via camera calibration. And these reconstructed point clouds involve to rotation, outliers, occlusions and other perturbations, which is enough to evaluate the performance of any registration algorithms
Comparable performance on a spatial memory task in data collected in the lab and online
No full textBrief description of the record: Our previous research highlighted a systematic bias in a spatial memory task, with participants correctly detecting object movements in the same direction as the perspective shift, whilst misjudging the direction of object movements if those were in the opposite direction to the perspective shift. The aim of the current study was to investigate if the introduction of perspective shifts results in systematic biases in object location estimations. To do so, we asked participants to encode the position of an object in a virtual room and to then estimate the object’s position following a perspective shift. In addition, by manipulating memory load (perception and memory condition) we investigated if the bias in object position estimates results from systematic distortions introduced in spatial memory. Overall, our results show that participants make systematic errors in estimating object positions in the same direction as the perspective shift. This bias was present in both the memory and the perception condition. We propose that the systematic bias in the same direction as the perspective shift is driven by difficulties in understanding the perspective shifts that may lead participants to use an egocentric representation of object positions as an anchor when estimating the object location following a perspective shift, thereby giving rise to a systematic shift in errors in the same direction as the perspective shift
Fair resource allocation algorithms (MLF-DRS, FFMRA, H-FFFMRA): The output and source codes for the experiments
No full textIn this thesis, we have used randomly generated workloads to evaluate the performance of proposed fair resource allocation algorithms (MLF-DRS, FFMRA, H-FFFMRA). The produced workloads have been directly fed into the simulation environment, considering different metrics such as resource allocation, utilization, and fairness. A sample CSV file has been provided that illustrates how the raw data generated by the mathematical models look (workload_0). This is a sample generated workload based on a single task with various resource demands in time-series experiments. The generated workload is used to determine allocation with respect to other tasks, submitted to the system
Graphic Medicine Exhibited: Alberda PhD Thesis Dataset
Get PDF’Graphic Medicine Exhibited’ employs an integrated mixed-methods methodology in order to answer its research questions and meet the aims and objectives for this study. The thesis uses a blended narrative inquiry and grounded theory methodology to create interviews and analyse their findings in chapters three and four. This analysis provides further insights into graphic medicine exhibition characteristics and barriers explored in chapters one and two. In addition to the methodological contributions of each chapter’s analyses, the research uses mixed theoretical and empirical methods through the interdisciplinary application of narrative inquiry and creative practice frameworks. The dataset here includes images from the exhibition analysed and interviews with curators and visitors
The role of memory and perspective shifts in systematic biases during object location estimation
No full textOnline data collection offers a wide range of benefits including access to larger and more diverse populations, together with a reduction in the experiment cycle. Here we compare performance in a spatial memory task, in which participants had to estimate object locations following viewpoint shifts, using data from a controlled lab-based setting and from an unsupervised online sample. We found that the data collected in a conventional laboratory setting and those collected online produced very similar results, although the online data was more variable with standard errors being about 10% larger than those of the data collected in the lab. Overall, our findings suggest that spatial memory studies using static images can be successfully carried out online with unsupervised samples. However, given the higher variability of the online data, it is recommended that the sample size should be increased by at least 25% in the data collected online, as this would allow to achieve similar standard errors to those obtained in the la
Environmental DNA as a non-invasive sampling tool to detect the spawning distribution of European anadromous shads (Alosa spp.)
No full text1. The shads Alosa alosa and Alosa fallax (Alosa spp.) have high conservation designations in Europe due to the vulnerability of their populations to human disturbances. In particular, river impoundments block their upstream migration, preventing access to spawning areas. Knowledge on the spatial extent of their spawning is important for informing conservation and river management plans.
2. Determining the spatial extent of Alosa spp. spawning is challenging. They enter rivers over a two to three-month period and the species potentially migrate different distances upstream. Capture and handling can be problematic and spawning events generally occur at night. Thus, assessing their spatial extent of spawning could utilise non-invasive sampling tools, such as environmental DNA
(eDNA).
3. An eDNA assay for Alosa spp. was successfully developed, based on the Cytochrome c Oxidase Subunit I gene segment and quantitative polymerase 82 chain reaction (qPCR). Application in spring 2017 to the River Teme (River Severn catchment, Western England) revealed high sensitivity in both laboratory and field trials. Field data indicated spawning between May and June, with migrants mainly restricted to areas downstream of the final impoundment.
4. eDNA can thus be utilised as a non-invasive sampling tool to determine the freshwater distribution of these fishes in Europe, enhancing their conservation at local and regional scales
Environmental factors are stronger predictors of primate species' distributions than basic biological traits
Get PDFUnderstanding the neutral, biological and environmental processes driving species distributions is valuable in informing conservation efforts because it will help us predict how species will respond to changes in environmental conditions. Environmental processes affect species differently according to their biological traits, which determine how they interact with their environment. Therefore, functional, trait-based modelling approaches are considered important for predicting distributions and species responses to change but even for data-rich primate communities our understanding of the relationships between traits and environmental conditions is limited. Here we use a large-scale, high-resolution dataset of African diurnal primate distributions, biological traits and environmental conditions to investigate the role of biological traits and environmental trait filtering in primate distributions. We collected data from published sources for 354 sites, and 14 genera with 57 species across Sub-Saharan Africa. We then combined a three-table ordination method, RLQ, with the Fourth Corner approach to test relationships between environmental variables and biological traits and used a mapping approach to visually assess patterning in primate genus and species' distributions. We found no significant relationships between any groups of environmental variables and biological traits, despite a clear role of environmental filtering in driving genus and species' distributions. The most important environmental driver of species distributions was temperature seasonality, followed by rainfall. We conclude that the relative flexibility of many primate genera means that not any one particular set of traits drives their species-environment associations, despite the clear role of such associations in their distribution patterns
Tipping points in lowland agricultural landscapes
No full textThis is an archive of research data associated with the project 'Tipping points in lowland agricultural landscapes'. This work was supported by NERC grant reference number: NE/P007716/1 as part of the Valuing Nature Programme (www.valuing-nature.net)