A novel multi-perspective imaging platform (M-PIP) for phenotyping soybean root crowns in the field increases throughput and separation ability of genotype root properties

Abstract

A b s t r a c t 15 Background: Root crown phenotyping has linked root properties to shoot mass, nutrient uptake, and 16 yield in the field, which increases the understanding of soil resource acquisition and presents 17 opportunities for breeding. The original methods using manual measurements have been largely 18 supplanted by image-based approaches. However, most image-based systems have been limited to one 19 or two perspectives and rely on segmentation from grayscale images. An efficient high-throughput root 20 crown phenotyping system is introduced that takes images from five perspectives simultaneously, 21 constituting the Multi-Perspective Imaging Platform (M-PIP). A segmentation procedure using the 22 Expectation-Maximization Gaussian Mixture Model (EM-GMM) algorithm was developed to 23 distinguish plant root pixels from background pixels in color images and using hardware acceleration 24 (CPU and GPU). Phenes were extracted using MatLab scripts. Placement of excavated root crowns for 25 image acquisition was standardized and is ergonomic. The M-PIP was tested on 24 soybean [Glycine 26 max (L.) Merr.] cultivars released between 1930 and 2005 . 27 Results: Relative to previous reports of imaging throughput, this system provides greater throughput 28 with sustained rates of 1.66 root crowns min -1 . The EM-GMM segmentation algorithm with hardware 29 acceleration was able to segment images in 10 s, faster than previous methods, and the output images 30 were consistently better connected with less loss of fine detail. Image-based phenes had similar 31 heritabilities as manual measures with the greatest effect sizes observed for Maximum Radius and Fine 32 Radius Frequency. Correlations were also noted, especially among the manual Complexity score and 33 phenes such as number of roots and Total Root Length. Averaging phenes across perspectives 34 3 generally increased heritability, and no single perspective consistently performed better than others. 35Angle-based phenes, Fineness Index, Maximum Width, Holes, Solidity and Width-to-Depth Ratio were 36 the most sensitive to perspective with decreased correlations among perspectives. 37Conclusion: The substantial heritabilities measured for many phenes suggest that they are potentially 38 useful for breeding. Multiple perspectives together often produced the greatest heritabilities, and no 39 single perspective consistently performed better than others. Thus, as illustrated here for soybean, 40 multiple perspectives may be beneficial for root crown phenotyping systems. This system can 41 contribute to breeding efforts that incorporate under-utilized root phenotypes to increase food security 42 and sustainability. 43

RhizoVision Crown: An Integrated Hardware and Software Platform for Root Crown Phenotyping

“…Over the past few years, the throughput, reliability, and standardization of root crown phenotyping has increased using digital imaging and image-based analysis software, such as DIRT [ 37 ], REST [ 32 ], and M-PIP [ 38 ]. Minimizing cost, increasing throughput, and improving reliability are key demands for developing high-throughput root phenotyping platforms.…”

“…In conclusion, RhizoVision Crown is a cost-effective and high-throughput platform that has the potential to increase access to technologies for root crown phenotyping. The platform builds upon previous platforms [ 29 , 32 , 37 , 38 ] by optimizing image acquisition using a backlight and the barcode option, using custom imaging software designed for phenotyping, and using image analysis software with a simple graphical interface designed for batch processing. All software are free and ready-to-use on Windows 10.…”

“…The Root Estimator for Shovelomics Traits (REST) platform included an imaging “tent” to provide more uniform lighting, a DSLR consumer camera controlled using the manufacturer's software, and a MATLAB executable for image analysis but does not extract root length [ 32 ]. The Multi-Perspective Imaging Platform (M-PIP) includes five point-and-shoot cameras along a 90 ° arc in an imaging box, command line camera control software for Linux, and MATLAB scripts for image analysis [ 38 ]. These platforms have advanced the field of root crown phenotyping, but advances can still be made to increase access to these technologies and to optimize imaging, image analysis, and data processing.…”

Crop Improvement from Phenotyping Roots: Highlights Reveal Expanding Opportunities

“…Root number, growth, and angle were measured in various species and soil conditions [62][63][64][65]. Digital imaging of root traits, root estimator for shovelomics traits, and the multiperspective imaging platform [66][67][68] standardize and increase the root measurements captured from multiple images of the same root crown. Both coring and shovelomics offer substantial opportunities when used with postprocessing to discover or validate traits.…”

Cost-effective, high-throughput phenotyping system for 3D reconstruction of fruit form

“…Publicly available methods for 2D phenotyping of plants and plant organs have increased in recent decades to support high quality analysis of leaves, roots, shoots, stems, tubers, and fruits (6, 20–27). Computer vision has shown great potential to quantify external fruit quality and 2D imaging has been successfully implemented to measure the shape and size of fruits such as strawberries (4, 12), apples (5), carrot (6, 14), mangoes (28), and many others.…”