An official website of the United States government
Plant counting is a critical aspect of crop management, providing farmers with valuable insights into seed germination success and within-field variation in crop population density, both of which are key indicators of crop yield and quality. Recent advancements in Unmanned Aerial System (UAS) technology, coupled with deep learning techniques, have facilitated the development of automated plant counting methods. Various computer vision models based on UAS images are available for detecting and classifying crop plants. However, their accuracy relies largely on the availability of substantial manually labeled training datasets. The objective of this study was to develop a robust corn counting model by developing and integrating an automatic image annotation framework. This study used high-spatial-resolution images collected with a DJI Mavic Pro 2 at the V2–V4 growth stage of corn plants from a field in Wooster, Ohio. The automated image annotation process involved extracting corn rows and applying image enhancement techniques to automatically annotate images as either corn or non-corn, resulting in 80% accuracy in identifying corn plants. The accuracy of corn stand identification was further improved by training four deep learning (DL) models, including InceptionV3, VGG16, VGG19, and Vision Transformer (ViT), with annotated images across various datasets. Notably, VGG16 outperformed the other three models, achieving an F1 score of 0.955. When the corn counts were compared to ground truth data across five test regions, VGG achieved an R2 of 0.94 and an RMSE of 9.95. The integration of an automated image annotation process into the training of the DL models provided notable benefits in terms of model scaling and consistency. The developed framework can efficiently manage large-scale data generation, streamlining the process for the rapid development and deployment of corn counting DL models.
more »
« less
Three-dimensional biphase fabric estimation from 2D images by deep learning
Abstract A pruned VGG19 model subjected to Axial Coronal Sagittal (ACS) convolutions and a custom VGG16 model are benchmarked to predict 3D fabric descriptors from a set of 2D images. The data used for training and testing are extracted from a set of 600 3D biphase microstructures created numerically. Fabric descriptors calculated from the 3D microstructures constitute the ground truth, while the input data are obtained by slicing the 3D microstructures in each direction of space at regular intervals. The computational cost to train the custom ACS-VGG19 model increases linearly withp(the number of images extracted in each direction of space), and increasingpdoes not improve the performance of the model - or only does so marginally. The best performing ACS-VGG19 model provides a MAPE of 2 to 5% for the means of aggregate size, aspect ratios and solidity, but cannot be used to estimate orientations. The custom VGG16 yields a MAPE of 2% or less for the means of aggregate size, distance to nearest neighbor, aspect ratios and solidity. The MAPE is less than 3% for the mean roundness, and in the range of 5-7% for the aggregate volume fraction and the mean diagonal components of the orientation matrix. Increasingpimproves the performance of the custom VGG16 model, but becomes cost ineffective beyond 3 images per direction. For both models, the aggregate volume fraction is predicted with less accuracy than higher order descriptors, which is attributed to the bias given by the loss function towards highly-correlated descriptors. Both models perform better to predict means than standard deviations, which are noisy quantities. The custom VGG16 model performs better than the pruned version of the ACS-VGG19 model, likely because it contains 3 times (p= 1) to 28 times (p= 10) less parameters than the ACS-VGG19 model, allowing better and faster cnvergence, with less data. The custom VGG16 model predicts the second and third invariants of the orientation matrix with a MAPE of 2.8% and 8.9%, respectively, which suggests that the model can predict orientation descriptors regardless of the orientation of the input images.
more »
« less
- Award ID(s):
- 2416344
- PAR ID:
- 10547728
- Publisher / Repository:
- Springer Nature
- Date Published:
- Journal Name:
- Scientific Reports
- Volume:
- 14
- Issue:
- 1
- ISSN:
- 2045-2322
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Hambleton, J. P. (Ed.)Soil particles that have been deposited through water or air generally align their largest projected surface area normal to the depositional direction, which generates a cross-anisotropic fabric of granular soils. Researchers have used both two-dimensional (2D) and three-dimensional (3D) images to determine scalar fabric parameters of granular soils, including void ratio, coordination number, and average branch vector length. This study aims to evaluate the accuracy and effectiveness of 2D images to characterize fabric in 3D soils based on scalar parameters. The X-ray computed tomography (X-ray CT) is used to reconstruct the 3D volumetric images of three air-pluviated sand specimens, including crushed limestone, Griffin sand, and glass beads. Then, six slices are obtained by vertically cutting the 3D volumetric image in an angle increment of 30 degrees. The 3D and 2D images are analyzed to determine scalar fabric parameters. The results show that coordination numbers and average branch vector lengths computed from 2D images underestimate these values in 3D granular soils. The void ratios computed from 2D images vary a large range depending on slicing directions, which cannot provide reliable fabric characterizations for 3D granular soils.more » « less
-
Abstract Manufacturing custom three-dimensional (3D) carbon functional materials is of utmost importance for applications ranging from electronics and energy devices to medicine, and beyond. In lieu of viable eco-friendly synthesis pathways, conventional methods of carbon growth involve energy-intensive processes with inherent limitations of substrate compatibility. The yearning to produce complex structures, with ultra-high aspect ratios, further impedes the quest for eco-friendly and scalable paths toward 3D carbon-based materials patterning. Here, we demonstrate a facile process for carbon 3D printing at room temperature, using low-power visible light and a metal-free catalyst. Within seconds to minutes, this one-step photocatalytic growth yields rod-shaped microstructures with aspect ratios up to ~500 and diameters below 10 μm. The approach enables the rapid patterning of centimeter-size arrays of rods with tunable height and pitch, and of custom complex 3D structures. The patterned structures exhibit appealing luminescence properties and ohmic behavior, with great potential for optoelectronics and sensing applications, including those interfacing with biological systems.more » « less
-
none (Ed.)An obstacle to developing a general mechanical framework for magma mush is the emergence and complexity of a crystal fabric. To illuminate the conditions that produce a crystal fabric we performed time-dependent numerical simulations using a Computational-Fluid-Dynamics and Discrete-Element-Method (CFD-DEM) model in three dimensions. The specific focus was on the role of shear strain in the creation of a preferential orientation of crystals in mush. CFD-DEM method allows for the simultaneous coupling and frictional interactions of melt and crystals undergoing shear strain. The crystal shapes are represented using spheroids (either oblate or prolate). Simulations consist in imposing a compression stress (pressure) and a simple shear to a dense suspension of crystals in a viscous liquid, and monitoring the evolution of the orientation and strength of the shape fabrics. We ran a series of simulations by varying the size and aspect ratio of the particles. We considered samples in which all the solids have the same volume and shape, and cases including size and aspect ratio distributions. Results show that the strength of the shape fabric and the angle between the crystal preferential orientation and the compression plane both increase with the shear strain up to steady state values, which are primarily controlled by the aspect ratio of the particles. The stronger the aspect ratio, the greater the magnitude of the preferential orientation and the lower its angle relative to the compression plane. When introducing a distribution in the size of the crystals, we observed a decrease in the strength of the shape fabric and an increase in the angle between the preferential orientation of the crystals and the compression plane compared with samples composed of crystals having the same shape and size. Similarly, the distribution in the aspect ratio further decreases the strength of the shape fabric and increases the angle between the preferential orientation of the solids and the compression plane. Finally, we employed an alternative approach to quantify the amount of foliation and lineation and show that the samples always display a stronger foliation than lineation, although the shear strain increases both the foliation and the lineation.more » « less
-
Fabric draping, which is referred to as the process of forming of textile reinforcements over a 3D mold, is a critical stage in composites manufacturing since it determines the fiber orientation that affects subsequent infusion and curing processes and the resulting structural performance. The goal of this study is to predict the fabric deformation during the draping process and develop in-depth understanding of fabric deformation through an architecture-based discrete Finite Element Analysis (FEA). A new, efficient discrete fabric modeling approach is proposed by representing textile architecture using virtual fiber tows modeled as Timoshenko beams and connected by the springs and dashpots at the intersections of the interlaced tows. Both picture frame and cantilever beam bending tests were carried out to characterize input model parameters. The predictive capability of the proposed modeling approach is demonstrated by predicting the deformation and shear angles of a fabric subject to hemisphere draping. Key deformation modes, including bending and shearing, are successfully captured using the proposed model. The development of the virtual fiber tow model provides an efficient method to illustrate individual tow deformation during draping while achieving computational efficiency in large-scale fabric draping simulations. Discrete fabric architecture and the inter-tow interactions are considered in the proposed model, promoting a deep understanding of fiber tow deformation modes and their contribution to the overall fabric deformation responses.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1038/s41598-024-59554-x
