Skip to content

How to Scan Plants with Plant Imager 2.1Link

Info

The Plant Imager 2.1 refers to the version fixed issues with the world reference frame.

Get ReadyLink

  1. Open a terminal - press Ctrl + Alt + t.
  2. Activate the environment 
    conda activate plant_imager
  3. Power the Sony RX0 camera - press the small black button.
  4. Connect the controller to the camera Wi-Fi (this should happen automatically).
  5. Turn on the Plant Imager unit.

Center the PlantLink

You must know where the center defined in your configuration is in the real world, then place the plant there.

  1. Open the scan configuration you intend to use (e.g. /data/ROMI/configs/scan.toml).

  2. Check the key parameters under [ScanPath] and [ScanPath.kwargs]: 

    [ScanPath]
class_name = "Circle"   # circular path

[ScanPath.kwargs]
center_x = 375   # X coordinate of the center
center_y = 375   # Y coordinate of the center
z        = 100   # Camera height (higher = lower camera)

Important

Keep center_x and center_y unchanged. You may adjust z as long as it stays <120.

  1. Move the arm to the center 

    romi_run_task ScannerToCenter /tmp --config /data/ROMI/configs/scan.toml

  2. When the arm stops, place your plant directly under it.

Run a Quick Test ScanLink

A short scan confirms that the plant is correctly positioned and fully visible.

  1. Set n_points to 4 in the same scan.toml (the other values stay the same): 
    [ScanPath.kwargs]
center_x = 375 # x-coordinate of the center
center_y = 375 # y-coordinate of the center
z = 100        # height of the camera, higher values lower the camera
radius = 350   # distance to the center / plant
n_points = 4  # number of images to take around the plant

Note

Do not modify radius unless you have a specific reason.

  1. Navigate to your acquisition folder and launch the test scan:
cd /path/to/my_database   # probably /data/ROMI/Romi_****/acquisitions
romi_run_task Scan test_4_views --config /data/ROMI/configs/scan.toml --no-auth

  1. Inspect the images Open the test_4_views/images folder in a file browser and verify that each picture captures the plant clearly.

  2. If the view is unsatisfactory, reposition the plant and repeat steps 2‑3 until you are happy with the results.

Information

The system has been tested up to center_x = 384, center_y = 384, and radius = 382

Scan a Plant DatasetLink

When you are ready to acquire a full dataset:

  1. Confirm the final scan parameters (radius and n_points) in scan.toml. Example for a dense acquisition:
[ScanPath.kwargs]
center_x = 375
center_y = 375
z = 100
radius = 350
n_points = 40   # more images for higher detail
  1. Add the plant’s biological metadata under [Scan.metadata.object] (replace the example values with your own):
[Scan.metadata.object]
species = "Arabidopsis thaliana"
ecotype = "Col0"
seed_stock = "186AV L5"
plant_id = "AL4_col0_158"
growth_environment = "C2_Lyon"
growth_conditions = "Long day"
treatment = "none"
DAG = 48
sample = "whole plant"
experiment_id = "AL4"

  1. Place the plant in the imager.

  2. Run the acquisition, naming the dataset (e.g. my_awesome_plant_007):

cd /path/to/my_database   # e.g. /data/ROMI/Romi_****/acquisitions
romi_run_task Scan my_awesome_plant_007 --config /data/ROMI/configs/scan.toml --no-auth

The task will capture the specified number of images around the plant.

Shut Down SafelyLink

  1. Stop the camera to prevent battery drain.
  2. Power off the Plant Imager unit.

Quick‑Reference ChecklistLink

Step Action
1 Open terminal, activate plant_imager, power camera, connect Wi‑Fi, turn on imager
2 Verify center_x, center_y, z; run ScannerToCenter; place plant
3 Set n_points = 4; run test_4_views; validate images
4 Adjust radius/n_points; fill metadata; run final Scan task
5 Stop camera, power down imager

Follow this guide whenever you need a reliable, repeatable plant scan with Plant Imager 2.1.