How ToLink
End usersLink
To get detailed instructions on how to perform tasks like data acquisition, reconstruction and quantification, follows these instructions:
AcquisitionsLink
- How to Scan Plants with Plant Imager 2.1
- How to Scan Plants with Plant Imager 3
- How to Scan Virtual Plants with the Virtual Plant Imager
Reconstructions & QuantificationLink
- How to Fix Outdated Reconstruction Pipeline Configuration
- How to Fix World Frame for Older Scan Datasets
- How to Choose Linear Coefficients and Threshold Parameters for Plant‑Mask Generation
- How to Reconstruct Plants with the Geometric Pipeline
- How to Reconstruct Plants with the ML Pipeline
- How to Reconstruct Virtual Plants with the Geometric Pipeline
Exploratory AnalysisLink
- How to Explore Reconstruction and Quantification with the Plant 3D Explorer
- How to extract phyllotactic angle and internode measurements from a PlantDB database
Tasks detailsLink
To get a deeper understanding of what each task does, we made some notebooks.
- Colmap - From RGB images to camera poses
- Undistorted - Undistort RGB images using camera model
- Masks - From RGB images to masks
- Voxels - From mask images and camera coordinates to volume
- PointCloud - From carved volume to point cloud
- TriangleMesh - From point cloud to triangular mesh
- CurveSkeleton - From triangular mesh to curved skeleton
- TreeGraph - From curved skeleton to tree graph
- AnglesAndInternodes - Extract fruits angles and internodes distance from a tree graph
You have to download them and execute them by calling jupyter notebook in the conda environment with the plant-3d-vision module installed.