A miniaturized phenotyping platform for individual plants using multi-view stereo 3D reconstruction

Abstract

Plant phenotyping is essential in plant breeding and management. High-throughput data acquisition and automatic phenotypes extraction are common concerns in plant phenotyping. Despite the development of phenotyping platforms and the realization of high-throughput three-dimensional (3D) data acquisition in tall plants, such as maize, handling small-size plants with complex structural features remains a challenge. This study developed a miniaturized shoot phenotyping platform MVS-Pheno V2 focusing on low plant shoots. The platform is an improvement of MVS-Pheno V1 and was developed based on multi-view stereo 3D reconstruction. It has the following four components: Hardware, wireless communication and control, data acquisition system, and data processing system. The hardware sets the rotation on top of the platform, separating plants to be static while rotating. A novel local network was established to realize wireless communication and control; thus, preventing cable twining. The data processing system was developed to calibrate point clouds and extract phenotypes, including plant height, leaf area, projected area, shoot volume, and compactness. This study used three cultivars of wheat shoots at four growth stages to test the performance of the platform. The mean absolute percentage error of point cloud calibration was 0.585%. The squared correlation coefficient R2 was 0.9991, 0.9949, and 0.9693 for plant height, leaf length, and leaf width, respectively. The root mean squared error (RMSE) was 0.6996, 0.4531, and 0.1174 cm for plant height, leaf length, and leaf width. The MVS-Pheno V2 platform provides an alternative solution for high-throughput phenotyping of low individual plants and is especially suitable for shoot architecture-related plant breeding and management studies.

A Synthetic Review of Various Dimensions of Non-Destructive Plant Stress Phenotyping

“…(i) Surface traits can be obtained using laser scanners (such as Light Detection and Ranging) and photogrammetry (such as stereo cameras) techniques. Laser scanners are used to obtain the spatial 3D "point clouds" through scanning, while photogrammetry uses stereovision 2D images to reconstruct 3D models [107][108][109]. (ii) Internal traits are often used to detect undersurface information unseen in ordinary ways.…”

Three-dimensional branch segmentation and phenotype extraction of maize tassel based on deep learning

“…At maize tassel loose pollen stage,the MVS-Pheno [29], an easy-to-assemble automated acquisition device with a supplemental light system, was used for image acquisition of maize tassels.…”

Using high-throughput phenotype platform MVS-Pheno to reconstruct the 3D morphological structure of wheat

“…In terms of plant 3D modelling, there are four main categories: rule-based methods ( Hidayat et al 2020 ; Cieslak et al 2022 ; Yang et al 2022 ), 3D digitizer-based methods ( Wang et al 2018 ; Chenxi et al 2022 ), image-based methods ( Lou et al 2014 ; Wu et al 2018 ; Das Choudhury et al 2020 ; Lu et al 2021 ; Wu et al 2022 ) and laser scanning-based methods ( Ma et al 2019 ; Sun and Wang 2019 ; Jin et al 2021 ; Sun et al 2021 ). Among them, the rule-based method can realize the 3D growth modelling of plants by specifying rules and parameterization.…”

“…This method uses structure from motion (SFM) ( Bolles et al 1987 ) to estimate the camera’s attitude information and reconstruct the sparse point cloud, then uses multiple view stereo (MVS) ( Goesele et al 2007 ) to estimate the dense point cloud. This method generates high-efficiency data acquisition with high point cloud density, and the plant high-throughput phenotyping platform developed based on this technology also has the advantages of low cost and easy deployment ( Wu et al 2022 ).…”