Related: Just the other day I used USGS 3DEP LiDAR data + Claude Code to get a sense for the number of trees on my property. Diffing terrain map and canopy map gives tree elevation. It was a fun project to explore, primarily because I set CC loose and said "here is the bounding box of my property, pad it by 50 feet and then go absolutely nuts against government datasets gathering as much open data as you can" - it figured out the rest. Dug into soil maps, historical satellite imagery, and lidar data.

Here are the visuals re: trees - https://i.imgur.com/R0W4q4O.png

The USGS Lidar data is a treasure trove, I use it a lot at work.

What did you do to actually count trees? Even from aerial Lidar it can be a bit finicky for closed canopies.

It's very rudimentary: smooth the canopy and find the local maxima above a base height. It's really just identifying the tallest points.

Here is the first pass, https://i.imgur.com/f7Gpxmm.png, it under counted and also even counted my house as a tree, lol.

> counted my house as a tree

Even the more sophisticated algorithms pretty much always do this ;-)

You are probably not interested in taking this further, but you could give the Li tree filter a try: https://pdal.org/en/stable/stages/filters.litree.html

But getting perfect segmentation is basically impossible.

Fascinating work and inspiring application of the underlying DINOv3 image segmentation model!

The blog post and paper [1] describe a promising approach to solving related problems at previously impossible scale and quality: I am currently exploring methods to better represent seasonal land cover changes that would improve wind power generation forecasting and this paper provides a great starting point.

I hope DINOv3 can inspire more work like this - and I would encourage any curious mind to play with that model! I was amazed by its capability to distinguish between fine object details. For example, in a photo of a bicycle, the patch embeddings cleanly separated the background from the individual spokes of the wheel.

[1] https://arxiv.org/abs/2603.06382

I gave a talk about the paper in our internal journal club recently (we work on similar problems, usually using stereo imagery though).

It's a nice piece of work. I especially like the sections on data cleaning and registration, as that seemed to have been one of the limiting factors of the previous approaches.

I am sceptical about how accurately you can predict heights for specific trees from mono-images, but I think for cases where you just need to be right on average (e.g. biomass estimation, fuel load estimates) it's a great approach.

why does meta map canopy heights?

I think they were buying carbon offsets at some point and trying to validate that the countries and organizations that were selling the carbon offset were not cutting down those trees, effectively profiting twice.

Presumably the smart ones just sell their promise-not-to-cut-down-my-forest multiple times. Laundered through completely trustworthy NGOs, so nothing can actually be audited properly.

It's from FAIR, i.e. their fundamental research arm.

Maybe there are some ulterior motives, but they do also just do a little bit of "feel-good" research.

This was also in collaboration with the World Resources Institute and the University of Maryland, so it's not a 100% facebook project.

This is really cool, I wonder how old the satellite data they used is, it’s a bit unclear

From the paper:

> CHMv2 is derived from single-date imagery, where the acquisition process selects the best available image within a target period (2017 -2020). This limits the direct use of the released CHMv2 data for attributing canopy height to a specified year of interest. To support change applications, we provide the image acquisition date associated with each prediction in the dataset metadata.

So generally a few years out of date, but the dataset is transparent about when each image was taken.

This is an important question.

The tree outside of house is not 9 feet tall per. I have a 2 story house and it easily towers 10 feet higher than my house.

Additionally, there are several Royal Palms that are close to 50ft and they show as being only 15 feet.

You could look it up in the metadata file:

> We additionally release a global GeoTIFF of input image acquisition date, where pixel values encode year minus 2000 (e.g., 18.25 indicates April 2018)

That being said, I am sceptical on how accurate mono-depth models can be on a single tree basis. I would probably trust them to do large scale biomass estimates, but probably not single tree height assessments.