Visualisation and Annotation of 360 Degree Imagery

Provides a customisable R 'shiny' app for immersively visualising, mapping and annotating panospheric (360 degree) imagery. The flexible interface allows annotation of any geocoded images using up to 4 user specified drop-down menus. The app uses 'leaflet' to render maps that display the geo-locations of images and Panellum < https://pannellum.org/>, a lightweight panorama viewer for the web, to render images in virtual 360 degree viewing mode. Key functions include the ability to draw on & export parts of 360 images for downstream applications. Users can also draw polygons and points on map imagery related to the panoramic images and export them for further analysis. Downstream applications include using annotations to train Artificial Intelligence/Machine Learning (AI/ML) models and geospatial modelling and analysis of camera based survey data.

{pannotator}

pannotator

The Panospheric Image Annotator in R (pannotator) software package provides an easy-to-use interface for visualising 360 degree camera images on satellite imagery and annotating the images with data selected from user-defined drop-down menus. It is designed for use in ecological and biogeographical research but can be used to extract data from any spatially explicit 360 degree camera imagery.

The current app includes:

• a modular workspace with Mapping, Image, Annotation, and Annotation Table panels
• persistent Settings that can sit below the workspace or open in a left/right drawer
• editable lookup and username .csv tables in Settings
• project YAML export for relaunching with run_app(projectSettingsFile = ...)
• ExifTool status/install helpers in Settings
• annotation exports that include a sourcekmz field

For a fuller walkthrough of setup, configuration, annotation workflows, and export options, see the package vignette.

Installation

Minimum Requirements

To use this package, please ensure your system meets the following minimum requirements:

• R version: 4.4.0 or higher

• RStudio version: 2024.04.2+764 or higher

• Shiny version: 1.9.1 or higher

Additionally, ensure that all necessary system dependencies are installed for optimal performance.

Below is some code to help ensure all dependencies are met:

The software makes extensive use of ExifTool by Phil Harvey (Exiftool.org). To make installation of ExifTool accessible in R there is a package exiftoolr that you must install by running the code below.

# First check if you have exiftoolr installed
check_for_package <- system.file(package = "exiftoolr")
print(check_for_package)

# If not run the following code
if (check_for_package == "") {
  print("exiftoolr package not found .....installing now")
  install.packages("exiftoolr")
} else {
  print("exiftoolr package is already installed")
}

Now that you have installed exiftoolr we can check to make sure that ExifTool is on your system. You can also check and install ExifTool from the System Dependencies section in the app Settings panel.

library(exiftoolr)  
check_for_ExifTool <- exiftoolr::exif_version(quiet = TRUE)

# Install ExifTool if not found
if (exists("check_for_ExifTool")) {
  print("ExifTool found on system")
  exiftoolr::exif_version()
} else {
  print("ExifTool not found ... installing now")   
  exiftoolr::install_exiftool()
}

Running the Package

To run the application use the following code.

library(pannotator)

options(shiny.port = httpuv::randomPort(), shiny.launch.browser = .rs.invokeShinyWindowExternal, shiny.maxRequestSize = 9000 * 1024^2)

run_app()

# Optional: launch with a project-specific YAML file exported from Settings
# run_app(projectSettingsFile = "C:/path/to/project.yml")

By default run_app() uses the regular user settings created by the package. If you export a project YAML from the Settings panel you can pass it back into run_app(projectSettingsFile = ...) to reopen the app with project-specific layout, lookup, and map/image settings.

Help Vignette

If you want help you can find it using the following code:

vignette('pannotator', package = 'pannotator')