Phil Bartie

Neighbourhoods and health: a GIS approach to measuring community resource accessibility

Objective: Recent studies suggest an association between the contextual attributes of neighbourhoods and the health status of residents. However, there has been a scarcity of studies that have directly measured the material characteristics of neighbourhoods theorised to have an impact on health and health inequalities. This paper describes the development of an innovative methodology to measure geographical access to a range of community resources that have been empirically linked to health. Geographical information systems (GIS) were applied to develop precise measures of community resource accessibility for small areas at a national scale. Design: Locational access to shopping, education, recreation, and health facilities was established for all 38 350 census meshblocks across New Zealand. Using GIS, distance measures were calculated from the population weighted centroid of each meshblock to 16 specific types of facilities theorised as potentially health related. From these data, indices of community resource accessibility for all New Zealand neighbourhoods were constructed. Results: Clear regional variations in geographical accessibility to community resources exist across the country, particularly between urban and rural areas of New Zealand. For example, the average travel time to the nearest food shop ranged from less than one minute to more than 244 minutes. Noticeable differences were also apparent between neighbourhoods within urban areas. Conclusions: Recent advances in GIS and computing capacity have made it feasible to directly measure access to health related community resources at the neighbourhood level. The construction of access indices for specific community resources will enable health researchers to examine with greater precision, variations in the material characteristics of neighbourhoods and the pathways through which neighbourhoods impact on specific health outcomes.

Development of a speech-based augmented reality system to support exploration of cityscape

When people explore new environments they often use landmarks as reference points to help navigate and orientate themselves. This research paper examines how spatial datasets can be used to build a system for use in an urban environment which functions as a city guide, announcing Features of Interest (FoI) as they become visible to the user (not just proximal), as the user moves freely around the city. Visibility calculations for the FoIs were pre-calculated based on a digital surface model derived from LIDAR (Light Detection and Ranging) data. The results were stored in a textbased relational database management system (RDBMS) for rapid retrieval. All interaction between the user and the system was via a speech-based interface, allowing the user to record and request further information on any of the announced FoI. A prototype system, called Edinburgh Augmented Reality System (EARS) , was designed, implemented and field tested in order to assess the effectiveness of these ideas. The application proved to be an innovating, ‘non-invasive’ approach to augmenting the user’s reality

Advancing visibility modelling algorithms for urban environments

This paper presents a number of methods for calculating the visibility of landmark features in the urban context. While current visibility modelling techniques establish which regions may be viewed from a location, it is sometimes necessary to quantify how much of a designated feature is visible from its surroundings. This is particularly relevant in the field of Location Based Services, where information is currently filtered using proximity and syntax matching, but could include more advanced egocentric and contextual filtering capabilities if feature visibility modelling was available. This research presents a method to establish the visible extents of landmarks in an urban environment, through the development of a number of visual metrics. These metrics are tested in both synthetic and real world trials.

An Egocentric Urban Viewshed: A Method for Landmark Visibility Mapping for Pedestrian Location Based Services

A variety of information can be provided to pedestrians using location based services in support of tasks such as wayfinding. Typically current location aware systems use proximity to filter databases for contextual information. We show that a filter based on the visibility of features is a useful additional capability made possible through the use of digital surface models. A number of visibility metrics are suggested for adoption by a location based service, to provide quantitative visibility information so that items of interest may be ranked according to a meaningful priority. Real world experiences validate the usefulness of these metrics, and a number of improvements are suggested.

An empirical model for estimating census unit population exposure in areas lacking air quality monitoring

This study presents the methods and results of part of the HAPiNZ (Health and Air Pollution in New Zealand) study. A part of this project was to produce accurate measures of pollution exposure for the entire population of New Zealand living in urban areas. Suitable data are limited in most parts of New Zealand with some areas having no monitoring at all. As a result, this project has developed an empirical model to estimate annual exposure values for the whole country down to the census area unit level. This uses surrogate emission indicators and meteorological variables. Data sources used include census data on domestic heating, industrial emissions estimates, vehicle kilometres travelled and meteorological measurements. These were used to calculate annual exposure estimates and were then compared to monitored data for the areas where monitoring data were available. Results show a good association between the model estimates and the monitored data, enabling advanced health effects assessments for the countrys entire urban population.

Incorporating vegetation into visual exposure modelling in urban environments

Visual exposure modelling establishes the extent to which a nominated feature may be seen from a specified location. The advent of high-resolution light detection and ranging (LiDAR)-sourced elevation models has enabled visual exposure modelling to be applied in urban regions, for example, to calculate the field of view occupied by a landmark building when observed from a nearby street. Currently, visual exposure models access a single surface elevation model to establish the lines of sight (LoSs) between the observer and the landmark feature. This is a cause for concern in vegetated areas where trees are represented as solid protrusions in the surface model totally blocking the LoSs. Additionally, the observers elevation, as read from the surface model, would be incorrectly set to the tree top height in those regions. The research presented here overcomes these issues by introducing a new visual exposure model, which accesses a bare earth terrain model, to establish the observers true elevation even when passing through vegetated regions, a surface model for the city profile and an additional vegetation map. Where there is a difference between terrain and surface elevations, the vegetation map is consulted. In vegetated areas the LoS is permitted to continue its journey, either passing under the canopy with clear views or partially through it depending on foliage density, otherwise the LoS is terminated. This approach enables landmark visual exposure to be modelled more realistically, with consideration given to urban trees. The models improvements are demonstrated through a number of real-world trials and compared to current visual exposure methods.

From the Virtual to the RealWorld: Referring to Objects in Real-World Spatial Scenes

Predicting the success of referring expressions (RE) is vital for real-world applications such as navigation systems. Traditionally, research has focused on studying Referring Expression Generation (REG) in virtual, controlled environments. In this paper, we describe a novel study of spatial references from real scenes rather than virtual. First, we investigate how humans describe objects in open, uncontrolled scenarios and compare our findings to those reported in virtual environments. We show that REs in real-world scenarios differ significantly to those in virtual worlds. Second, we propose a novel approach to quantifying image complexity when complete annotations are not present (e.g. due to poor object recognition capabitlities), and third, we present a model for success prediction of REs for objects in real scenes. Finally, we discuss implications for Natural Language Generation (NLG) systems and future directions.

Route Ahead Visibility Mapping: A method to model how far ahead a motorist may view a designated route

Abstract Please click here to download the map associated with this article. We examine how visibility analysis may be used to calculate the extent of the route ahead visible to a motorist. To do this a Digital Surface Model sourced from a LiDAR dataset is used, which includes building and vegetation profiles as well as topography. Once the route has been designated the visibility along the path is modelled, in the direction of travel, from the drivers viewpoint. The visibility analysis considers all surrounding objects during the calculation, but reports only how far ahead the route may be viewed, and not the total landscape area visible. The map accompanying the article shows the results from using this method for a route in Christchurch, New Zealand, from which we identify a number of interesting results along the route. This method may be used such that navigational devices customise announcements to the most opportune times when the drivers workload is minimal and decision points are in view, rather than at predefined distances. It could also be used to notify the driver of the safest overtaking locations or those areas of limited visibility.

Identifying related landmark tags in urban scenes using spatial and semantic clustering

Abstract There is considerable interest in developing landmark saliency models as a basis for describing urban landscapes, and in constructing wayfinding instructions, for text and spoken dialogue based systems. The challenge lies in knowing the truthfulness of such models; is what the model considers salient the same as what is perceived by the user? The method developed in this research identifies related annotated tags supplied from a web based experiment in which users were asked to tag the most salient features on urban images for the purposes of navigation and exploration. The tag collections may be used to rank landmark popularity in each scene, but the challenge is in determining which tags relate to the same object (e.g. tags relating to a particular cafe). Existing clustering techniques did not perform well for this task, and it was therefore necessary to develop a new spatial-semantic clustering method which considered the proximity of nearby tags and the similarity of their label content. The annotation similarity was initially calculated using trigrams in conjunction with a synonym list, generating a set of networks formed from the links between related tags. These networks were used to build related word lists encapsulating conceptual connections (e.g. church tower related to clock) so that during a secondary pass of the data, related network segments could be merged. This approach gives interesting insight into the partonomic relationships between the constituent parts of landmarks and the range and frequency of terms used to describe them.

Understanding Information Requirements in "Text Only" Pedestrian Wayfinding Systems

Information that enables an urban pedestrian to get from A to B can come in many forms though maps are generally preferred. However, given the cognitive load associated with map reading, and the desire to make discrete use of mobile technologies, there is increasing interest in systems that deliver wayfinding information solely by means of georeferenced spoken utterances that essentially leave the user “technology free.” As a critical prior step, this paper examines the optimal delivery of such georeferenced text based instructions in anticipation of their spoken utterance. We identify the factors governing the content, location of instruction and frequency of delivery of text instructions such that a pedestrian can confidently follow a prescribed route, without reference to a map. We report on street level experiments in which pedestrians followed a sequence of text instructions delivered at key points along a set of routes. In examining instructions that are easy to follow, we compare landmark based instructions with street name based instructions. Results show that a landmark based approach is preferred because it is easier to assimilate (not because it is faster). Analysis also revealed that some degree of redundancy in the instructions is required in order to bring “comfort” to the user’s progress. There still remains the challenge of modeling the saliency of landmarks, knowing what is the most efficient set of instructions, and how to vary the frequency of instruction according to the complexity of the route. The paper concludes by identifying a set of design heuristics useful in the design of text based instructions for wayfinding.