multiple questions

Section for FAQs for i-Tree Canopy

Moderators: i-Tree Support, i-Tree Team

Post Reply
Baharak
Posts: 3
Joined: Fri Jun 26, 2020 8:39 am

multiple questions

Post by Baharak » Fri Jun 26, 2020 9:01 am

hi,
I am using i-tree canopy for the assessment of canopy in a small town in Ontario. During this process, some questions came to my mind and I would be pleased if you could guide me through these questions:
1- is the classification based on the spectral signature of the features(similar to remote sensing approaches)?
2- in identifying the classes for each sample point, do I have to specifically focus on the coverage of the cross-bar or I can consider the whole view?
3- can you tell me what is the area under the cross-bar?
4- does i-tree consider textures and patterns or it is focused just on pixels?
5- what is your suggestion for occasions where cross-bar is located on mixed coverages, especially on edges of two coverages?
6- can I add new classes or modify the current classes later during identifying sample points?

thanks in advance for your help
Jason.Henning
i-Tree Team
Posts: 216
Joined: Fri Dec 21, 2012 7:45 am

Re: multiple questions

Post by Jason.Henning » Fri Jun 26, 2020 10:08 am

Thanks for the questions. Please see my answers below.
Thanks,
Jason

1- is the classification based on the spectral signature of the features(similar to remote sensing approaches)?
There is no automated classification. The classification is wholly based on the users interpretation of what is shown in the image under the cross-hairs.

2- in identifying the classes for each sample point, do I have to specifically focus on the coverage of the cross-bar or I can consider the whole view?
Users should focused on the cover at the exact center of the cross-hair.

3- can you tell me what is the area under the cross-bar?
The area varies based on how much you are zoomed in and it is not actually used or recorded by the application. The percentage covers are calculated based on what percentage of your points fall in each cover-type. Please see this documentation also available under the "Help" menu in i-Tree Canopy for additional details, https://canopy.itreetools.org/references.

4- does i-tree consider textures and patterns or it is focused just on pixels?
The application is not considering anything shown in the images beyond how user's classify each individual point.

5- what is your suggestion for occasions where cross-bar is located on mixed coverages, especially on edges of two coverages?
User's should zoom in as far as possible and consider the cover at the exact center of the cross-hairs. If it is really not possible to discern you should pick one covertype and the next time you see a similar issue select the other.

6- can I add new classes or modify the current classes later during identifying sample points?
This is not possible as the tool is using random sampling and there is no way to reconcile new categories or changes with already classified points. It is important to carefully consider project goals before set-up so that you make sure you design the project and the classification scheme to meet those goals.
A member of the i-Tree Team
Baharak
Posts: 3
Joined: Fri Jun 26, 2020 8:39 am

Re: multiple questions

Post by Baharak » Fri Jun 26, 2020 10:12 am

thank you very much for your answers
Post Reply