Trail Conditions and Community Use: Utilizing Geospatial Video to Guide the Adoption of a Spatial-Temporal Trail Audit Tool (STAT)

Physical activity (PA), associated with all-cause mortality, morbidity, and healthcare costs, improves vitamin D absorption, immune response, and stress when completed outdoors. Rural communities, which experience PA inequities, rely on trails to meet PA guidelines. However, current trail audit methods could be more efficient and accurate, which geospatial video may support. Therefore, the study purpose was (1) to identify and adopt validated instruments for trail audit evaluations using geospatial video and a composite score and (2) to determine if geospatial video and a composite score motivate (influence the decision to use) specific trail selection among current trail users. Phase 1 used a mixed-method exploratory sequential core design using qualitative data, then quantitative data for the development of the Spatial-temporal Trail Audit Tool (STAT). Geospatial videos of two Northeast Ohio trails were collected using a bicycle-mounted spatial video camera and video analysis software. The creation of STAT was integrated from Neighborhood Environment Walkability Scale (NEWS), Walk Score, and Path Environment Audit Tool (PEAT) audit tools based on four constructs: trail accessibility, conditions, amenities, and safety. Scoring was determined by three independent reviewers. Phase 2 included a mixed-method convergent core design to test the applicability of STAT for trail participant motivation. STAT has 20 items in 4 content areas computing a composite score and was found to increase trail quality and motivation for use. STAT can evaluate trails for PA using geospatial video and a composite score which may spur PA through increased motivation to select and use trails.

Applying Spatial Video Geonarratives and Physiological Measurements to Explore Perceived Safety in Baton Rouge, Louisiana

Spatial crime analysis, together with perceived (crime) safety analysis have tremendously benefitted from Geographic Information Science (GISc) and the application of geospatial technology. This research study discusses a novel methodological approach to document the use of emerging geospatial technologies to explore perceived urban safety from the lenses of fear of crime or crime perception in the city of Baton Rouge, USA. The mixed techniques include a survey, spatial video geonarrative (SVG) in the field with study participants, and the extraction of moments of stress (MOS) from biosensing wristbands. This study enrolled 46 participants who completed geonarratives and MOS detection. A subset of 10 of these geonarratives are presented here. Each participant was driven in a car equipped with audio recording and spatial video along a predefined route while wearing the Empatica E4 wristbands to measure three physiological variables, all of them linked by timestamp. The results show differences in the participants’ sentiments (positive or negative) and MOS in the field based on gender. These mixed-methods are encouraging for finding relationships between actual crime occurrences and the community perceived fear of crime in urban areas.

Exploring convolutional neural networks and spatial video for on-the-ground mapping in informal settlements

Background The health burden in developing world informal settlements often coincides with a lack of spatial data that could be used to guide intervention strategies. Spatial video (SV) has proven to be a useful tool to collect environmental and social data at a granular scale, though the effort required to turn these spatially encoded video frames into maps limits sustainability and scalability. In this paper we explore the use of convolution neural networks (CNN) to solve this problem by automatically identifying disease related environmental risks in a series of SV collected from Haiti. Our objective is to determine the potential of machine learning in health risk mapping for these environments by assessing the challenges faced in adequately training the required classification models. Results We show that SV can be a suitable source for automatically identifying and extracting health risk features using machine learning. While well-defined objects such as drains, buckets, tires and animals can be efficiently classified, more amorphous masses such as trash or standing water are difficult to classify. Our results further show that variations in the number of image frames selected, the image resolution, and combinations of these can be used to improve the overall model performance. Conclusion Machine learning in combination with spatial video can be used to automatically identify environmental risks associated with common health problems in informal settlements, though there are likely to be variations in the type of data needed for training based on location. Success based on the risk type being identified are also likely to vary geographically. However, we are confident in identifying a series of best practices for data collection, model training and performance in these settings. We also discuss the next step of testing these findings in other environments, and how adding in the simultaneously collected geographic data could be used to create an automatic health risk mapping tool.

Where’s The Crime? Exploring Divergences Between Call Data and Perceptions of Local Crime

Scholars typically use calls to the police to study crime patterning; however, crime reporting may be systematic across space. Using spatial video and geonarrative methodology, we investigate the overlap between perceived crime hot spots among 35 neighbourhood insiders (police officers, ex-offenders and residents) and hot spots gleaned from call data. In general, perceptual hot spots diverge from call data, but in particular, a corner store emerges as a perceptual hot spot across all groups, but not in call data. We use our data to explore the microgeographic dynamics of this ‘hidden hot spot’. We find that the corner store is relatively isolated, with few occupied residences around it and participants avoiding it. In addition, our geonarratives suggest that the store lacks adequate guardianship. We argue that mixed methodological approaches such as these are useful for understanding discrepancies between measures as well as the situational and environmental dynamics of problem places.