On the relationship between multi-view data capturing and quality of rendered virtual view

Sx, Liu; An, Ping; Zy, Zhang; Zhang, Q; Shen, L-Q; Jiang, G-Y

doi:10.1179/136821909x12476507838352

Cited by 9 publications

(6 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A rail track along the curved frame allows for fine adjustment of the camera tilt and pan direction without compromising stability (Figure 2d). Cameras are then arranged along the curved frame to achieve a sampling of bigger and smaller root morphology that satisfy the Nyquist theorem to prevent aliasing (Liu et al, 2009). In the case of maize roots, more cameras are concentrated to image the root crown with high amounts of small occluded roots.…”

Section: Methodsmentioning

confidence: 99%

DIRT/3D: 3D root phenotyping for field grown maize (Zea mays)

Liu

Barrow

Hanlon

et al. 2020

Preprint

View full text Add to dashboard Cite

AbstractThe development of crops with deeper roots holds substantial promise to mitigate the consequences of climate change. Deeper roots are an essential factor to improve water uptake as a way to enhance crop resilience to drought, to increase nitrogen capture to reduce fertilizer inputs and to increase carbon sequestration from the atmosphere to improve soil organic fertility. A major obstacle to achieving these improvements is the bottleneck in the high-throughput phenotyping of field-grown roots. We address this bottleneck with DIRT/3D, a newly developed image-based 3D root phenotyping platform, which measures 16 architecture traits in mature maize root systems. DIRT/3D computed reliably all 16 traits on a test panel of 12 contrasting maize genotypes. The analysis of the maize panel validated traits such as distance between whorls, and the number, angle, and diameters of crown and brace roots. Overall, we observed a coefficient of determination of r2>0.84 and a high broad-sense heritability of for all traits. Means of the 16 traits and a newly developed descriptor to characterize a complete root architecture distinguished all genotypes. DIRT/3D is a step towards automated quantification of highly occluded maize root systems in the field and will support breeders and root biologists to improved carbon sequestration and food security in the face of the adverse effects of climate change.

show abstract

Section: Methodsmentioning

confidence: 99%

DIRT/3D: 3D root phenotyping for field grown maize (Zea mays)

Liu

Barrow

Hanlon

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…For each sample, approximately 360 images with an image resolution of 5,472 × 3,648 were taken. The number of images was chosen after evaluating the scanning geometry of this 3D scanner following the method described in Liu et al (2009). We obtained an upper bound of the number of images needed to avoid under-sampling of the multi-view scanning setup.…”

Section: 21mentioning

confidence: 99%

Comparison of open‐source three‐dimensional reconstruction pipelines for maize‐root phenotyping

Liu

Bonelli

Pietrzyk

et al. 2023

The Plant Phenome Journal

View full text Add to dashboard Cite

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.

show abstract

“…For the rendering quality quantify and evaluation, Liu et al [13] derived a proper image quantity for rendering and a 3D surface describing the connection between multi-view data capturing and quality of the rendered view by analysing the spectral bounds of the scene for the camera model. They also achieved a balance between camera arrangement and the number of virtual views.…”

Section: Introductionmentioning

confidence: 99%

Camera Position Optimization Using the Effective Area In Light Field Rendering

2014

Proceedings of International Conference on Internet Multimedia Computing and Service

View full text Add to dashboard Cite

In light field rendering, images captured by cameras are used to render virtual views. The changes in camera positions lead to different rendering quality. To improve rendering quality, it is necessary to optimize the camera position. This paper proposes a mathematical model based on a definition called effective area (EA) to rearrange cameras. EA connects camera position with the rendering error so as to quantify the suitability of a camera position for a given arrangement of virtual views. Using the EA, we present an optimization algorithm for camera position to yield the optimal quality. Simulate results demonstrate that EA can be used to quantify the rendering quality. We also show the optimization results of different arrangement of virtual views.

show abstract

On the relationship between multi-view data capturing and quality of rendered virtual view

Cited by 9 publications

References 17 publications

DIRT/3D: 3D root phenotyping for field grown maize (Zea mays)

DIRT/3D: 3D root phenotyping for field grown maize (Zea mays)

Comparison of open‐source three‐dimensional reconstruction pipelines for maize‐root phenotyping

Camera Position Optimization Using the Effective Area In Light Field Rendering

Contact Info

Product

Resources

About