Close
0%
0%

Virtual Forest

track the seasons through tele-presence

Public Chat
Similar projects worth following
The Virtual Forest project gives everyone the opportunity to track the seasons through tele-presence (VR). It allows for leaf peeping from a distance and visualizes ecological processes for those who find access to green spaces difficult. Virtual Forest allows you to immersively track the seasons using 360 degree photography of the experimental forest plot at Gontrode, Belgium. Media can be viewed immersively using a Google Cardboard or similar on the project website:

https://virtualforest.io

The images gathered support further research efforts in understanding the timing of changes to the canopy structure. But, above all, I hope Virtual Forest will inspire people to venture outdoors and explore a forest in real life.

Technical details

The Virtual Forest project combines a Rico Theta 360 camera and a raspberry pi with a custom outdoor housing into a rugged 360 (outdoor) time-lapse camera. The housing is made of cheap standard PVC fittings, sitting on top of a garden fence post. The optics are covered by a glass lamp shade to provide optimal transmission and limited deformation (acrylic globes can be used as well).  The camera is mounted in a custom 3D printed PLA setup. Black PLA is used to limit internal reflections during dusk and dawn of any status lights. Constructed and placed in the forest the camera looks like a garden lamp (see figure).

An ethernet cable, which runs to a nearby powered hub, serves as the internet and power connection. The setup has a ground wire for surge protection due to voltage spikes from any nearby lightning strikes over the long wire run.

Data is acquired by a simple cron job on a raspberry pi, and uploaded to a cloud provider (various protocols can be used but I rely on RClone to support most available data services). Currently images are taken every half hour, which is plenty due to the relatively slow varying nature of vegetation growth.

The most recent image is visualized on a simple website using a VR A-frame setup for immersive interactive viewing.

Multiple use cases

The Virtual Forest project grew out of a personal interest in using off-the-shelve consumer hardware with interesting properties (360 field of view) within the context of scientific outreach and data acquisition. The data can be used to present a unique (VR) perspective to changes in the forest. For example, the light environment has an important function in the understory of the forest, but it is hard to visualize how this might change over the course of a day. Using data from the project a 360 video can illustrate this process, and provide ways to digitally explore the forest environment. Other visualizations include scrolling through a whole year using for example Panomoments.



More so, being involved in the US based PhenoCam network for my research I knew the strength of simple cameras as powerful (visual) tools to assess the state and model vegetation development [1 - 2] and health [3]. Aside from the very visual scientific outreach component the project serves research efforts into seasonal monitoring of vegetation growth and health. Curated data is released from time to time and deposited in permanent storage location (i.e. Zenodo.org). The latest release contains curated data for the year 2017 at Harvard forest [4]. Additional data releases are foreseen, but priority is given at MSc and PhD students using the most recent data in their research.

Useful links

Project repository (installation instructions, software, hardware)

- https://github.com/bluegreen-labs/photosphere


Ricoh Theta 360 camera lineup

- https://theta360.com/en/


Media coverage

- https://www.raspberrypi.com/news/virtual-forest/

- https://www.vice.com/en/article/bmvpq8/virtual-forest


References

1. Hufkens, K., Melaas, E. K., Mann, M. L., Foster, T., Ceballos, F., Robles, M. and Kramer, B. 2019. Monitoring crop phenology using a smartphone based near-surface remote sensing approach. – Agricultural and Forest Meteorology 265: 327–337.

2. Hufkens, K., Keenan, T. F., Flanagan, L. B., Scott, R. L., Bernacchi, C. J., Joo, E., Brunsell, N. A., Verfaillie, J. and Richardson, A. D. 2016. Productivity of North American grasslands is increased under future climate scenarios despite rising aridity. – Nature Climate Change <https://doi.org/10.1038/nclimate2942>.

3. Hufkens, K., Friedl, M. A., Keenan, T. F., Sonnentag, O., Bailey, A., O’Keefe, J. and Richardson, A. D. 2012. Ecological impacts of a widespread frost event following early spring leaf-out. – Global Change Biology 18: 2365–2377.

4. Hufkens, Koen. (2021). Spherical forest phenology images - 2017 - Harvard Forest (v1.0) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.4899637

... Read more »

sla - 451.64 kB - 08/12/2022 at 14:23

Download

sla - 583.19 kB - 08/12/2022 at 14:23

Download

  • 1 × Ricoh Theta 360 camera Most models would work, newer ones have higher resolution images and are thermally more stable
  • 1 × raspberry pi 3B+ A 3B+ or newer is required to use the standard PoE hat
  • 1 × Raspberry pi PoE hat
  • 1 × 48V PoE injector
  • 1 × APC surge protector protects against power surges due to lightning on long ground wire runs

View all 12 components

  • Network issues streaming site

    Koen Hufkens09/03/2022 at 20:08 0 comments

    It seems that the project might be gathering some above average traffic to the website. I apologize for things being slow or not loading / properly refreshing. Since this is a side project I won't be bumping up traffic quota. Before I could rely on generous capacity of Harvard University to host the project. This is sadly not the case anymore (and the website is therefore a bit of a hack).

  • Virtual Forest

    Koen Hufkens08/12/2022 at 14:48 0 comments

    Hi everyone,


    I'm reviving a long running project of mine. Although this has received some exposure in the past I've updated the build construction and the location has changed as well. Historically the location of the Virtual Forest was Harvard Forest in Western Massachusetts, USA. As I've changed positions, and moved back to the EU, maintaining the camera wasn't possible anymore. Although the staff at Harvard Forest kept up with for a while after I left, in the end the camera was removed.

    Over the past couple of years, with COVID confining many people indoors with little access to green spaces for leisure or education it became obvious that the original idea needed to be revived (and was more pressing than ever).

    Halfway through the COVID pandemic I therefore put together a new build. The result of this build, as well as the interactive experience is described and linked to in this project. Enjoy.

    Cheers,

    Koen

View all 2 project logs

  • 1
    Hardware


    1. From a length of 10 cm PVC pipe cut roughly 60 - 70 cm of pipe. This will house the Raspberry pi and the Ricoh Theta housing will rest on top of this. Insert the wooden fence pole into the PVC pipe and mount permanently with some wood screws through the base of the PVC pipe. Leave plenty of room for the Raspberry pi and wiring. Cut the wooden post to length, with the top of the PVC pipe reaching your chin, and attach the ground anchor. At this point you can install the pole at the desired location, and run Cat5e cable to the location for PoE power (other power solutions are possible but not explored here). To avoid animals from entering the PVC housing it is advised to fill all holes with tin foil or plastic bags.

    2. Take the glass globe and set it in cardboard box with some rags, with the opening pointing upward. Place the PVC coupler over the base of the glass globe (this should fit - check the specs when ordering so that the mounting hole flange does not exceed the inner diameter of the coupler). Squeeze (black) silicon mounting glue between the outer flange of the glass globe and the inner PVC coupler. Silicon glue can be a mess, to prevent drops falling in the glass globe you can fill it with toilet paper or rags. Leave the glue to set before removing the glass globe from the box (or the rags from the inside of the globe). To prevent any water ingress use the same silicon glue to seal any outside gaps between the glass globe and the PVC coupler.

    3. Print the attached STL files using (black) PLA (or other stiff material). Using 3mm screws mount the top and bottom half of the PLA construction. A small nodge must be made in the top of the PVC pipe to make the construction fit neatly in the PVC pipe. This ensures that when doing maintenance you retain the orientation of the camera. If not you will need to align all the images, which is not a trivial task. A description on how to approach this should you have misalignment is given here.

    4. Install all required software on a Raspberry pi 3B+ (or more recent). For software instructions see below.

    5. Using the 2.5 mm screws delivered with the PoE shield mount the Raspberry pi onto the base of the 3D printed construction. Attach a micro-usb cable to the pi and run the cable through the base plate of the 3D printed construction to the top.

    6. Use 3 mm set screw to set the height of the Ricoh Theta camera within the construction. Depending on the model you can vary this slightly. Hook up the camera and turn it on.

    At this point you should be done and the camera should take pictures on a regular interval.

  • 2
    Software


    On a fresh raspberry pi install clone the photosphere project, and run the install script. This will setup a cron job and put all scripts in the correct place. Nothing fancy here.

    git clone https://github.com/bluegreen-labs/photosphere.git
    sh install_photosphere.sh

    Change the `capture_sphere.sh` script in order to accommodate your own data storage solution (either local or remote). I used the rclone program to assign a remote drive (pcloud / onedrive / dropbox etc) where to store the images. I refer to the RClone webpage for the setup of such a remote location.

    https://github.com/bluegreen-labs/photosphere/blob/bc5c30d39588384626d564b43666abf1fd8baa18/capture_sphere.sh#L224

    Lines past 228 specify how I refresh a simple github page content to store the most recent image (which is a quick hack not to deal with cross site scripting requirements). You can adjust this to your own liking, or just delete if you do not need a live view.

    https://github.com/bluegreen-labs/photosphere/blob/bc5c30d39588384626d564b43666abf1fd8baa18/capture_sphere.sh#L228

View all instructions

Enjoy this project?

Share

Discussions

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates