John Gibson, Senior Geospatial Analyst
Watercourse monitoring has traditionally been conducted throughout urban New Zealand using the Watercourse Assessment Methodology: Infrastructure and Ecology version 2.0. While highly effective this technique is labour intensive and doesn’t deliver a comprehensive digital data set for post analysis and review. Working in partnership with Zealandia Consulting we saw an opportunity to refresh the assessment process using 360 video.
This blog summarises the technical processes used in gathering, integrating and disseminating data from stream surveys. This is based on processes trialled for the Raumanga Stream Survey (Whangarei) with Zealandia Ltd, with a view to future reuse of the processes. Our approach to data capture is summarised in the figure below.
The main outputs of this workflow is a database of observation data captured during field surveys, plus a collection of panoramic digital images (still images or videos) captured at the same time. The main tasks and learnings in each are summarised here.
Field capture
During a Streamwalk Survey, a customised ArcGIS Field App is used on a mobile phone to capture track data and observations, e.g., pipe infrastructure details. Captured data is transferred over cell connections and stored in the cloud-hosted ArcGIS Online system. A GoPro Max 360 panoramic camera is also used to continually record 3D imagery during the survey.
Initially, the GoPro 360 was used to capture panoramic video in short (8 minute) segments. The video could then be replayed on a desktop PC, allowing synchronised voice recordings as well as GPS tracks to match with a GIS map view. GoPro does not provide tools to extract still images from video at regular intervals; also image extraction removes the GPS location data from the images. Some customisation tools were created to extract still images from video, but these are reasonably time-consuming to use.
It is recommended that images are captured as timelapse still images at 2, 5 or 10 sec intervals on the GoPro Max. Whilst there are now several alternatives to the GoPro Max camera on the market, recent reviews suggest that GoPro Max remains the best option for panoramic image capture in the field.
Desktop processing
Desktop processing and data integration is done using ArcGIS. Extensions exist for ArcGIS Pro to allow panoramic imagery viewing from both Mapillary as well as using ESRI’s “Oriented Imagery Catalog” (OIC) tools. This allows map locations to be viewed alongside the relevant pano image view.
Panoramic images can be more widely shared using 3 methods :
Upload to Mapillary. Mapillary is a free image hosting and viewing platform owned by Meta. Any imagery uploaded is publicly shared & downloadable by others. Mapillary is like a open-source version of Google Streetview, with all Mapillary content being crowd-sourced.
Upload to cloud storage (such as Amazon S3). This is chargeable but inexpensive. This should be used for integration using OIC tools as above and can be used in public web viewers with ESRI’s “Experience Builder” software.
Upload to Google Streetview. Uploading is a similar process to using Mapillary. However, the imagery is not downloadable by others once uploaded.
Panoramic imagery can also be viewed in ArcGIS Pro, as shown above. This allows editing of map features and integration of data captured during field survey. In ArcGIS Pro, panoramic imagery can be viewed either using the Mapillary extension, or the Esri Oriented Imagery Catalog (OIC) extension – both are freely available.
Shared access to data
Data dissemination to clients and the public is normally by creation of technical reports with image figures. This can usefully be extended and improved by providing web apps for shared use, which can integrate much of the report data. Web apps can also include access to panoramic imagery and potentially also text reporting tools.
Initial work attempted to integrate a Mapillary viewer into a GIS web app. This feature was promoted by ESRI in 2022 as an addition to their Experience Builder software, for creating web apps. However, this does not currently (April 2023) seem to be an available option.
Suggested options include:
Use of ESRI Experience Builder software with the OIC extension to create a web app with embedded pano image viewer (see below).
Image above shows integration of panoramic imagery using the OIC extension in a Esri Experience Builder app.
Use of Mapillary, however this does not provide full data integration, as the map location is shown but cannot be customised.
Other alternatives would involve use of open-source software to integrate a GIS viewer with a pano viewer using the Mapillary API (programming interface).
In summary a data capture workflow has been fully prototyped and implemented successfully for the Raumanga Stream. Future work will include more continuous data capture, adding LIDAR-derived datasets to the project and further iterative development with the Esri Experience Builder and the OIC extension.
The Raumanga Stream view is publicly available on Mapillary.