What tree is this?

First, look at this google streeview:

How cool would it be, if I can generate a detection algorithm for tree boundary, like this:

And then, extract the LiDAR data from the scene

Classify it, and say –> “This is a Norway Maple”

41st Canadian Symposium On Remote Sensing. Call for Abstract: Special Session (AI in Remote Sensing)

Session Chairs: Dr. Connie Ko, Dr. Gunho Sohn. York University

The expansion in accessibility of remote sensing data combined with increases in data quantity and complexity of the incoming data has also increased. This challenge requires new theory, methods and system implementations in remote sensing data analytics. Recently, deep learning has been demonstrating an enormous success in visual data analytics. Deep learning is a non-representation learning model, which allows data to learn from its own representation. Due to its generalizability, this technology has been widely used in both computer science and remote sensing community. However, when applying these methods to outdoor natural environment and natural objects, there are still many problems to be solved, which include, but are not limited to: learning with small, noisy, out-of-distribution training data, domain and knowledge transfer, active, continual and fine-grained learning, integrating physical priors, and data fusions. This special session will bring an opportunity to discuss the current status of the AI technology adopted in the remote sensing research community and discuss its limitations and potential for directing our future research efforts.

How to Apply?
To submit your abstract, please visit https://csrs2020.exordo.com where you will find the 2020 ExOrdo abstract submission link. Note that you will be required to setup an account first.

Extended deadline for submission: Feb 28, 2020. Extended abstracts up to 400 words are invited.

The detail of the conference can be found https://crss-sct.ca/conferences/csrs-2020/. Please share widely with any students or researches that may be interested.  

Automatic Individual Tree Crown Delineation

Marker-controlled watershed algorithm is very simple and effective. LiDAR point cloud is converted into 2D canopy height model (raster model) and if you can imagine inverting the raster model. The depression of the surface model can be identified as tree top and the edge of the derived watershed boundary can be identified as tree crown boundary. The biggest challenges for this method is to overcome the assumption that one tree hold a single tree top and the fine the proper “scale” for segmenting the watershed. It’s sometimes managed by train-and-error approach for finding those parameters. I’m currently looking for ways to overcome these challenges.