Phenotiki: an open software and hardware platform for affordable and easy image‐based phenotyping of rosette‐shaped plants

Abstract

Phenotyping is important to understand plant biology, but current solutions are costly, not versatile or are difficult to deploy. To solve this problem, we present Phenotiki, an affordable system for plant phenotyping that, relying on off-the-shelf parts, provides an easy to install and maintain platform, offering an out-of-box experience for a well-established phenotyping need: imaging rosette-shaped plants. The accompanying software (with available source code) processes data originating from our device seamlessly and automatically. Our software relies on machine learning to devise robust algorithms, and includes an automated leaf count obtained from 2D images without the need of depth (3D). Our affordable device (~€200) can be deployed in growth chambers or greenhouse to acquire optical 2D images of approximately up to 60 adult Arabidopsis rosettes concurrently. Data from the device are processed remotely on a workstation or via a cloud application (based on CyVerse). In this paper, we present a proof-of-concept validation experiment on top-view images of 24 Arabidopsis plants in a combination of genotypes that has not been compared previously. Phenotypic analysis with respect to morphology, growth, color and leaf count has not been performed comprehensively before now. We confirm the findings of others on some of the extracted traits, showing that we can phenotype at reduced cost. We also perform extensive validations with external measurements and with higher fidelity equipment, and find no loss in statistical accuracy when we use the affordable setting that we propose. Device set-up instructions and analysis software are publicly available ( http://phenotiki.com).

Description of olive morphological parameters by using open access software

“…Finally, other commercial equipment or platforms which offer phenomics analysis were developed mostly for studying the growth of plants. Therefore, their aim was differed compared to our objective [46–49]. …”

“…For example, for the separation of fruits, leaves and endocarps from the image background, we are using the manual colour thresholding technique, which is commonly used in many applications in plants. However, this procedure can be less biased by the user, by using interactive segmentation methods or recent developed tools for such purposes [49, 59]. Moreover, the proposed methodology does not allow the user to interact within the workflow and calculate a parameter such as the determination of the apex point of a fruit which was attached to the pedicel.…”

“…This methodology, in addition to the identification of the wounded olive fruits, provided some basic morphological features only for the fruits such as height and perimeter. Finally, the needs of the plant research community has driven the emergence of various approaches for phenotyping the growth of plants by using either software or commercial equipment which one of their attributes is the determination of growth related measurements [46–49]. …”

Pheno‐Deep Counter: a unified and versatile deep learning architecture for leaf counting

“…Interestingly, intra-and inter-observer variation can also be used to assess algorithm performance. Based on the findings of Giuffrida et al (2018), inter-observer variation has a MSE of 0.81 [inexperienced annotators on a subset of Arabidopsis images used in Minervini et al (2017)]. Experienced and inexperienced annotators are within the AE1 leaf error range in about 90% of cases, whereas PhenoDC is within AE1 leaf error in about 80% of cases, thus bringing us closer to human-level performance.…”

“…They both evaluated their general method in plant images of wild‐type Arabidopsis based on an open dataset (Minervini et al ., ). However, the collection of such finely grained annotations is tedious and time‐consuming (Minervini et al ., 2015b; Minervini et al ., ), particularly when one must annotate data of significant diversity to account for large leaf variation, different imaging conditions, etc. In addition, these methods are very sensitive to how leaves are arranged (i.e.…”