Suxing Liu

Arate 3D reconstruction is essential for high-throughput plant phenotyping, particularly for studying complex structures such as root systems. While photog

Most computer vision‐ and machine learning‐based plant phenotyping systems compute traits such as shape and size rather than the color distribution of the plant surface, even though color can provide important insights into plant physiology. Therefore, we developed Speedy Measurement of Arabidopsis Rosette Traits (SMART), an open‐source plant phenotyping pipeline that analyzes a red–green–blue (RGB) top‐view image captured by any imaging device to compute color traits as well as shape and size. SMART combines a pretrained U2‐Net machine learning model and a color clustering method to segment plants from their background and compute basic morphological traits. SMART showed a good average accuracy of 95% for morphological traits using a public benchmark dataset. Uniquely, SMART also analyzes the color of plant surfaces by calculating a normalized color difference index and comparing plant surface colors with reference colors in the L*a*b* color space, which are converted from the RGB color space. The color difference index also showed good correlation with independent measurements of the chlorophyll fluorescence parameter F _v / F _m (maximum quantum yield of photosystem II) ( R ² > 0.71), chlorophyll content ( R ² > 0.73), and leaf temperature ( R ² > 0.76) in our experimental conditions. Therefore, we show that SMART is not only an affordable, open‐source tool for calculating morphological traits such as shape and size but also it is also useful for exploring relationships between color traits and physiological traits. SMART represents a promising new approach to low‐cost, high‐throughput phenotyping, thus benefiting the entire plant science community.

Accurate 3D reconstruction is essential for high-throughput plant phenotyping, particularly for studying complex structures such as root systems. While photogrammetry and Structure from Motion (SfM) techniques have become widely used for 3D root imaging, the camera settings used are often underreported in studies, and the impact of camera calibration on model accuracy remains largely underexplored in plant science. In this study, we systematically evaluate the effects of focus, aperture, exposure time, and gain settings on the quality of 3D root models made with a multi-camera scanning system. We show through a series of experiments that calibration significantly improves model quality, with focus misalignment and shallow depth of field (DoF) being the most important factors affecting reconstruction accuracy. Our results further show that proper calibration has a greater effect on reducing noise than filtering it during post-processing, emphasizing the importance of optimizing image acquisition rather than relying solely on computational corrections. This work improves the repeatability and accuracy of 3D root phenotyping by giving useful calibration guidelines. This leads to better trait quantification for use in crop research and plant breeding.

BioStudies – one package for all the data supporting a study

SUMMARY: The first draft of the Arabidopsis genome was released more than 20 years ago and despite intensive molecular research, more than 30% of Arabidopsis genes remained uncharacterized or without an assigned function. This is in part due to gene redundancy within gene families or the essential nature of genes, where their deletion results in lethality (i.e., the dark genome). High‐throughput plant phenotyping (HTPP) offers an automated and unbiased approach to characterize subtle or transient phenotypes resulting from gene redundancy or inducible gene silencing; however, access to commercial HTPP platforms remains limited. Here we describe the design and implementation of OPEN leaf, an open‐source phenotyping system with cloud connectivity and remote bilateral communication to facilitate data collection, sharing and processing. OPEN leaf, coupled with our SMART imaging processing pipeline was able to consistently document and quantify dynamic changes at the whole rosette level and leaf‐specific resolution when plants experienced changes in nutrient availability. Our data also demonstrate that VIS sensors remain underutilized and can be used in high‐throughput screens to identify and characterize previously unidentified phenotypes in a leaf‐specific time‐dependent manner. Moreover, the modular and open‐source design of OPEN leaf allows seamless integration of additional sensors based on users and experimental needs.

Understanding three‐dimensional (3D) root traits is essential to improve water uptake, increase nitrogen capture, and raise carbon sequestration from the atmosphere. However, quantifying 3D root traits by reconstructing 3D root models for deeper field‐grown roots remains a challenge due to the unknown tradeoff between 3D root‐model quality and 3D root‐trait accuracy. Therefore, we performed two computational experiments. We first compared the 3D model quality generated by five state‐of‐the‐art open‐source 3D model reconstruction pipelines on 12 contrasting genotypes of field‐grown maize roots. These pipelines included COLMAP, COLMAP+PMVS (Patch‐based Multi‐View Stereo), VisualSFM, Meshroom, and OpenMVG+MVE (Multi‐View Environment). The COLMAP pipeline achieved the best performance regarding 3D model quality versus computational time and image number needed. In the second test, we compared the accuracy of 3D root‐trait measurement generated by the Digital Imaging of Root Traits 3D pipeline (DIRT/3D) using COLMAP‐based 3D reconstruction with our current DIRT/3D pipeline that uses a VisualSFM‐based 3D reconstruction on the same dataset of 12 genotypes, with 5–10 replicates per genotype. The results revealed that (1) the average number of images needed to build a denser 3D model was reduced from 3000 to 3600 (DIRT/3D [VisualSFM‐based 3D reconstruction]) to around 360 for computational test 1, and around 600 for computational test 2 (DIRT/3D [COLMAP‐based 3D reconstruction]); (2) denser 3D models helped improve the accuracy of the 3D root‐trait measurement; (3) reducing the number of images can help resolve data storage problems. The updated DIRT/3D (COLMAP‐based 3D reconstruction) pipeline enables quicker image collection without compromising the accuracy of 3D root‐trait measurements.

Understanding three-dimensional (3D) root traits is essential to improve water uptake, increase nitrogen capture, and raise carbon sequestration from the atmosphere. However, quantifying 3D root traits by reconstructing 3D root models for deeper field-grown roots remains a challenge due to the unknown tradeoff between 3D root-model quality and 3D root-trait accuracy. Therefore, we performed two computational experiments. We first compared the 3D model quality generated by five state-of-the-art open-source 3D model reconstruction pipelines on 12 contrasting genotypes of field-grown maize roots. These pipelines included COLMAP, COLMAP+PMVS (Patch-based Multi-view Stereo), VisualSFM, Meshroom, and OpenMVG+MVE (Multi-View Environment). The COLMAP pipeline achieved the best performance regarding 3D model quality versus computational time and image number needed. Thus, in the second test, we compared the accuracy of 3D root-trait measurement generated by the Digital Imaging of Root Traits 3D pipeline (DIRT/3D) using COLMAP-based 3D reconstruction with our current DIRT/3D pipeline that uses a VisualSFM-based 3D reconstruction (Liu et al., 2021) on the same dataset of 12 genotypes, with 5~10 replicates per genotype. The results revealed that, 1) the average number of images needed to build a denser 3D model was reduced from 3000~3600 (DIRT/3D [VisualSFM-based 3D reconstruction]) to 300~600 (DIRT/3D [COLMAP-based 3D reconstruction]); 2) denser 3D models helped improve the accuracy of the 3D root-trait measurement; 3) reducing the number of images can help resolve data storage capacity problems. The updated DIRT/3D (COLMAP-based 3D reconstruction) pipeline enables quicker image collection without compromising the accuracy of 3D root-trait measurements.

Quantifying phenotypes of root-root interactions would allow a greater understanding of how plants react to belowground competition through plasticity of architectural traits. Past research has shown that plants will over proliferate roots in the presence of competition, leaving less resources to allocate above ground, negatively impacting shoot growth and yields [1]. Further evidence highlights a response to neighboring plants in the root architecture of Arabidposis thaliana, as individuals concentrate root mass towards their competitors [2]. To visualize and quantify root architecture plasticity involved in these root-root interactions in real soil, we developed a modified mesocosm system. Within the mesocosm box common bean (Phaseolus vulgaris) seeds were germinated 10 inches apart from each other. Mesh screens were placed on either side of each bean, in order to capture root growth towards each other and/or away from each other. Plants were harvested at the 7-week mark, when the root archetype of each individual was developed and prominent. To harvest our boxes three sides of the mesocosm box was removed and the soil was washed away. Images of each box was taken to produce 3D models. Improvements to the mesocosm system will be made to reliably extract root traits in real soil. These experiments will shine light on an understudied section of crop science and will allow farmers and researchers a better understanding of an otherwise unseen phenomena.

Understanding root traits is essential to improve water uptake, increase nitrogen capture and accelerate carbon sequestration from the atmosphere. High-throughput phenotyping to quantify root traits for deeper field-grown roots remains a challenge, however. Recently developed open-source methods use 3D reconstruction algorithms to build 3D models of plant roots from multiple 2D images and can extract root traits and phenotypes. Most of these methods rely on automated image orientation (Structure from Motion)[1] and dense image matching (Multiple View Stereo) algorithms to produce a 3D point cloud or mesh model from 2D images. Until now the performance of these methods when applied to field-grown roots has not been compared tested commonly used open-source pipelines on a test panel of twelve contrasting maize genotypes grown in real field conditions[2-6]. We compare the 3D point clouds produced in terms of number of points, computation time and model surface density. This comparison study provides insight into the performance of different open-source pipelines for maize root phenotyping and illuminates trade-offs between 3D model quality and performance cost for future high-throughput 3D root phenotyping.