Shoot Phenobox. The Phenobox is a open hardware phenotyping system (Czedik‐Eysenberg et al. 2018) for the assessment of shoot traits of small to medium-sized plants (not rosette). The 'PhenoBox' itself consists of an imaging chamber that autonomously captures and processes the shoot images from different perspectives. The hardware is complemented by 'PhenoPipe', an open source analytical framework for sample management and evaluation.
The standard Shoot Phenobox has the following external dimensions: 75.5 cm high, 50 cm wide and 82 cm (side) + 4 cm (door handle) long. It weighs approximately 15 kg. The Imaging area is suitable for plants with a maximum shoot size of approximately 20 cm width and 40 cm height. We are happy to discuss the manufacture of custom Shoot Phenoboxes with other dimensions (e.g. for plants with taller/wider canopies).
The camera can be adjusted in height to achieve a centered position on the shoot for imaging. Homogeneous illumination of the imaging area is provided by 7 LED panels, two on each side of the imaging axis (vertical) and three (horizontal) above the imaging compartment. Any camera compatible with gphoto2 can be used in the Phenobox; in the standard Phenobox, a Canon SLR camera (700D) is installed for easy operation and high image quality. Focus can be set automatically or manually (i.e. recommended setting for smaller shoots). The system is controlled by a Rasberry Pi 3 microcomputer.
Pots are positioned with a pot holder on a rotational stand; entry point is through the front door (standard) or custom side or top doors/lids. The included pot holder has two sides for square pot sizes with 4.5 or 5.5 cm edges and various heights (standard); custom pot holders (for round pots, etc.) can be manfactured to meet the required experimental conditions. Pots are labeled with QR codes for automatic plant registration in the PhenoPipe server system (see below). Label printer & software (suitable for setting up experimental schemes) included. The standard Phenobox comes with 6 preset rotation angles for imaging each plant shoot, ranging from 0° to 300° in 60° increments (absolute positions). To set custom rotation angles, see the GitHub documentation on the motor controller or discuss a different factory preset with us.
Please note that a web server and a WLAN connection are required to run the Phenobox - PhenoPipe imaging integration. Please ask your institution's IT administrator to help you set this up (only needed once; see Phenobox WLAN Setup / Server Configuration (below / on GitHub). Vienna Scientific will provide assistance in setting up the system. User must supply:
The code for operating the imaging routine of the Phenobox (via Raspberry Pi 3, installed inside the Phenobox electronic compartment), is factory deployed by Vienna Scientific.
Please contact us to discuss possible modifications (e.g. chamber dimensions, additional openings for plant access from the side or top, pot holder sizes required, rotation angles set, etc.) to the original Shoot Phenobox designs. We would be happy to engage in a joint research project extending the PhenoBox imaging capabilities.
Once the imaging pipeline is configured (see below), operation the Phenobox is straightforward and high throughput.
1) Start Rasberry Pi ("green button"; on right side of image), 2) Open door, place plant/pot (with QR code towards front); at the beginning: adjust camera position to the center of the shoot (height), set focus, 3) Start imaging routine (press "red button"), wait for "Blue" indicator lamp to flash (finished), 4) Repeat step 3 for all plants, and when finished, 5) turn off Rasberry Pi / LEDs (pressing the green button). Wait until turning off power until the green/blue lamps stopped flashing (i.e. Rasberry Pi shutdown completed).
Four indicator lamps (left side on image above) indicate potential imaging or server connection issues. See details on GitHub regarding the colour codes.
For possible connection of the Rasberry Pi, a central USB connector (black cover; next to the red button) has been added to the original Phenobox hardware design.
PhenoPipe Image Analysis. While VSI provides the Phenobox hardware, a matching image analysis pipeline was developed by Czedik-Eysenberg et al. (2018) and is available as open source on GitHub. https://github.com/Gregor-Mendel-Institute/PhenoBox-System. A recent, updated version of the server settings etc. to run the imaging routine based on the Phenobox is available and will be supplied with the assembled hardware - ensuring that your open source-driven phenotyping projects can start instantly.
For the original Shoot PhenoBox publication, please refer to Czedik‐Eysenberg et al. (2018). New Phytol. https://doi.org/10.1111/nph.15129. Source codes & Instructions are deposited on GitHub and requirements / functions of each server are described in brief below.
This is the central server application of the Phenobox shoot phenotyping system. It is used to create, store and manage phenotyping projects.
System requirements / set-up
In short, a key part of the PhenoPipe system is the IAP (Integrated Analysis Platform) server (Klukas, 2013); the actual image analysis is performed using IAP. Czedik‐Eysenberg et al. (2018) have integrated this analysis platform into an automated workflow to minimize user effort in performing multiple analyses on images. IAP analysis is always performed on all snapshots for a given timestamp.
IAP System Requirements / Setup
IAP Analysis pipelines
Researchers can upload their own IAP pipelines tuned for their experiments and have them available for selection in the web interface before starting an analysis at a given time. For information on creating and tuning IAP pipelines, please refer to the IAP user documentation (pdf). To create these pipelines, you need an IAP installation or you can use pipelines created by others.
In short, this server is used for post-processing of the IAP analysis (above). Users have the possibility to upload their own R scripts to this server, which can then be selected via the web interface to be applied to the data. This allows scripts to be run in an automated way on a server (with more computational power than a local computer).
System Requirements / Setup
No R installation is required because Czedik‐Eysenberg et al. (2018) set up Renjin to run the R scripts. This results in faster code execution than standard R, and the communication between the Java server code and the actual R scripts allows multiple scripts to run in parallel. More information on R scripting for the Shoot Phenobox pipeline.