Plant structure recognition
I’m developing a project that needs to make a model of branches of an apple tree.
I’m using Kinect v2.0 as computer vision.
In my application I need to create a structure model of the plant that is in front of the Kinect. Using Kinect I can make a 3D pointcloud of the plant in front of the camera.
Basically I should create a model with position of every visible branch, starting point and end point of the branch, orientation of the branch, and an estimation of the diameter of every branch.
It’s enough to make the model of the visible part, no problem if I can’t see hidden branches.
I would like to know if it’s possible to use Computree for this purpose, and if you can give me some tips how to do it.
Thanks a lot,
RE: Plant structure recognition - Added by Hackenberg Jan over 2 years ago
there is a SimpleTree plugin in Computree which I develop. In case with your clouds you can model the tree, you can get all the paramters you describe.
I though do not have experience with Kinect clouds. I see following problems:
- My plugin is developed for building Tree models of mature trees. Larger tress are in geneal more easy to model. Not their twigs, they have the same problems.
- My plugin is developed to use TLS clouds. I guess those are higher resoluted and more accurate than Kinect clouds.
- My plugin is developed for multiple scan mode. The modelling quality is also known to drop of for single scan design which has more occlusions.
If your cloud contains unexpected entries like NAN, it might not be computable at all.
This should say, you can try out, but there is no warranty.
You can get more information/an impression here:
RE: Plant structure recognition - Added by Heurtebize Julien over 2 years ago
RE: Plant structure recognition - Added by Roc Rom over 2 years ago
Thanks a lot to both of you for your quick and useful reply, I will ask you again in case of needs.