Objective Definition of Rosette Shape Variation Using a Combined Computer Vision and Data Mining Approach

Abstract

Computer-vision based measurements of phenotypic variation have implications for crop improvement and food security because they are intrinsically objective. It should be possible therefore to use such approaches to select robust genotypes. However, plants are morphologically complex and identification of meaningful traits from automatically acquired image data is not straightforward. Bespoke algorithms can be designed to capture and/or quantitate specific features but this approach is inflexible and is not generally applicable to a wide range of traits. In this paper, we have used industry-standard computer vision techniques to extract a wide range of features from images of genetically diverse Arabidopsis rosettes growing under non-stimulated conditions, and then used statistical analysis to identify those features that provide good discrimination between ecotypes. This analysis indicates that almost all the observed shape variation can be described by 5 principal components. We describe an easily implemented pipeline including image segmentation, feature extraction and statistical analysis. This pipeline provides a cost-effective and inherently scalable method to parameterise and analyse variation in rosette shape. The acquisition of images does not require any specialised equipment and the computer routines for image processing and data analysis have been implemented using open source software. Source code for data analysis is written using the R package. The equations to calculate image descriptors have been also provided.

Dissecting the Phenotypic Components of Crop Plant Growth and Drought Responses Based on High-Throughput Image Analysis

“…As reported here, our pipeline is capable of parallel processing of image data from multiple sensors and supports the extraction of a large number of relevant traits (Klukas et al, 2014). The number of traits, including image-based features and model-derived parameters, extracted from our pipeline greatly exceeds existing pipelines (Wang et al, 2009;Hartmann et al, 2011;De Vylder et al, 2012;Green et al, 2012;Paproki et al, 2012;Zhang et al, 2012;Camargo et al, 2014). We applied sophisticated methods to select a list of representative traits that are powerful in revealing descriptive phenotypic patterns of plants.…”

Arabidopsis phenotyping through geometric morphometrics

“…Chosen landmarks were present from early stages of infection to later stages and placed in regions that experience dramatic changes when infected [39]. Also, selected landmarks could take into account relevant morphological changes induced by stresses or distinctive phenotypes of different ecotypes such as relative shortening or lengthening of petioles and laminae or relative lateral displacement of leaves [7, 10, 11]. Moreover, the landmarks chosen are probably less prone to manual digitization error than, e.g., a landmark situated in the middle of the laminae or placed somewhere along the leaf’s contour.…”

“…However, the persistence of ad hoc descriptors [12, 13] and lack of a gold standard could give rise to reproducibility issues, because of different growing substrate-segmentation algorithm combinations. Moreover, different approaches sometimes give the same name to different parameters (e.g., “roundness” in ImageJ, [14] vs. [10]) or different names to the same parameter (e.g., “solidity” in [11] equals “compactness” in [7, 10] and “surface coverage” in [5]). The need to develop objective, mathematically, and statistically sound and more accurate shape descriptors in plants has been stressed in recent reviews on the topic [15–17].…”

Deep Phenotyping: Deep Learning for Temporal Phenotype/Genotype Classification

“…Previous work on phenotyping different genotypes (accession) have mostly used biologist specified, 'hand-crafted' image features such as number of leaves, leaf area, compactness, roundness, etc. [4,5,6,7,8]. These features are computed either manually or via custom image processing algorithms.…”