Senior Design Projects

ECS193 A/B Winter & Spring 2021

Tree size estimation app

Email **********
Heiner Lieth
Plant Sciences Department, UCDavis

Project's details

Tree size estimation app
In the plant sciences we frequently need to know the size of trees. While we can sometimes walk up to a tree with a tape measure or ruler and take some measurements, this is not always possible. Most notably mature trees tend to be so large that even the simple height measurement is a challenge. Old-fashioned tools do exist (sextants, rulers, tape measures,…) as do some expensive tools that use lasers. But their use is time-consuming. Many persons who need such information are generally satisfied with estimates; but it is relevant how approximate the estimate is. While knowing the height is interesting and important, there are other measures that are also important in various professions. For example: trunk diameter at chest height is important to forest managers and ecologists.
The challenge in this project is to use image processing directly in a mobile phone to identify a free-standing tree and calculate, in addition to tree height, the canopy width (ie the extent of the branches and leaves) and canopy volume, as well as trunk volume (ie how much lumber is in that trunk?). The expectation is that the app would run under AndroidOS or IOS and allow a forest or orchard manager to quickly assess the various parameters for many trees at particular locations, each identified with latitude and longitude. It is anticipated that a user might walk up to a tree to have the mobile device obtain the location coordinates with its built-in GPS sensor, then the user might back up to where an image of the entire tree can be taken (perhaps with panoramic mode in case the tree is very large) and image processing is then used to obtain the desired measures so quickly that one could measure many trees in quick successions as would be interesting to an orchard manager. The individual tree data would be stored in data files for further processing with other tools.

I would expect that the use could be tested within our Oak tree collection in the UCDavis Arboretum (near the Gazebo). The trees there are very large to where a person cannot really even see the entire tree from below for every tree. As such this may well be an extreme that the app cannot handle; that would be important to know. Another test would be the amazing Cork Oaks that we have on campus near South Hall and North Hall; for these it is generally possible to stand far enough back to get the entire specimen into a photo. Another test would be run in an orchard or other fruit tree planting near campus where the trees are yet young and can be measured with old-fashioned approaches to get ground truth comparison data.
The client expects these deliverables: (1) The app itself, as described above, running on the client's Android device; (2) ground-truth test data to confirm that the app is working as expected. (3) The computer code in ASCII files in whatever programming language the team chooses, with ample documentation within the code. (4) A written report where the code is an appendix and the testing is described.
Computer programming under AndroidOS and IOS, analytical geometry, image data processing
**********
30-60 min weekly or more
Open source project
Attachment N/A
No
Team members N/A
N/A
N/A