Observing plants across time and diverse scenes is critical in uncovering plant growth patterns.Classic methods often struggle to observe or measure plants against complex backgrounds and at different growth stages.Th...Observing plants across time and diverse scenes is critical in uncovering plant growth patterns.Classic methods often struggle to observe or measure plants against complex backgrounds and at different growth stages.This highlights the need for a universal approach capable of providing realistic plant visualizations across time and scene.Here,we introduce PlantGaussian,an approach for generating realistic three-dimensional(3D)visualization for plants across time and scenes.It marks one of the first applications of 3D Gaussian splatting techniques in plant science,achieving high-quality visualization across species and growth stages.By integrating the Segment Anything Model(SAM)and tracking algorithms,PlantGaussian overcomes the limitations of classic Gaussian reconstruction techniques in complex planting environments.A new mesh partitioning technique is employed to convert Gaussian rendering results into measurable plant meshes,offering a methodology for accurate 3D plant morphology phenotyping.To support this approach,PlantGaussian dataset is developed,which includes images of four crop species captured under multiple conditions and growth stages.Using only plant image sequences as input,it computes high-fidelity plant visualization models and 3D meshes for 3D plant morphological phenotyping.Visualization results indicate that most plant models achieve a Peak Signal-to-Noise Ratio(PSNR)exceeding 25,outperforming all models including the original 3D Gaussian Splatting and enhanced NeRF.The mesh results indicate an average relative error of 4%between the calculated values and the true measurements.As a generic 3D digital plant model,PlantGaussian will support expansion of plant phenotype databases,ecological research,and remote expert consultations.展开更多
Reliable prediction of the shale fracturing process is a challenging problem in exploiting deep shale oil and gas resources.Complex fracture networks need to be artificially created to employ deep shale oil and gas re...Reliable prediction of the shale fracturing process is a challenging problem in exploiting deep shale oil and gas resources.Complex fracture networks need to be artificially created to employ deep shale oil and gas reserves.Randomly distributed minerals and heterogeneities in shales significantly affect mechanical properties and fracturing behaviors in oil and gas exploitation.Describing the actual microstructure and associated heterogeneities in shales constitutes a significant challenge.The RFPA3D(rock failure process analysis parallel computing program)-based modeling approach is a promising numerical technique due to its unique capability to simulate the fracturing behavior of rocks.To improve traditional numerical technology and study crack propagation in shale on the microscopic scale,a combination of high-precision internal structure detection technology with the RFPA^(3D) numerical simulation method was developed to construct a real mineral structure-based modeling method.First,an improved digital image processing technique was developed to incorporate actual shale microstructures(focused ion beam scanning electron microscopy was used to capture shale microstructure images that reflect the distri-butions of different minerals)into the numerical model.Second,the effect of mineral inhomogeneity was considered by integrating the mineral statistical model obtained from the mineral nanoindentation experiments into the numerical model.By simulating a shale numerical model in which pyrite particles are wrapped by organic matter,the effects of shale microstructure and applied stress state on microcrack behavior and mechanical properties were investigated and analyzed.In this study,the effect of pyrite particles on fracture propagation was systematically analyzed and summarized for the first time.The results indicate that the distribution of minerals and initial defects dominated the fracture evolution and the failure mode.Cracks are generally initiated and propagated along the boundaries of hard mineral particles such as pyrite or in soft minerals such as organic matter.Locations with collections of hard minerals are more likely to produce complex fractures.This study provides a valuable method for un-derstanding the microfracture behavior of shales.展开更多
It is a key feature to embed 3D realistic sound effect in the future multimedia and virtual reality systems. Recent research on acoustics and psychoacoustics reveals the important cues for sound localization and sound...It is a key feature to embed 3D realistic sound effect in the future multimedia and virtual reality systems. Recent research on acoustics and psychoacoustics reveals the important cues for sound localization and sound perception. One promising approach to generate 3D realistic sound effect uses two earphones by simulating the sound waveforms from sound source to eardrum. This paper summarizes two methods for generating 3D realistic sound and points out their inherent drawbacks. To overcome these drawbacks we propose a simplified model to generate 3D realistic sound at any positions in the horizontal plane based on the results of sound perception and localization. Experimental results show that the model is correct and efficient.展开更多
基金supported by the Central Government’s Guidance Fund for Local Science and Technology Development(2024ZY-CGZY-19)。
文摘Observing plants across time and diverse scenes is critical in uncovering plant growth patterns.Classic methods often struggle to observe or measure plants against complex backgrounds and at different growth stages.This highlights the need for a universal approach capable of providing realistic plant visualizations across time and scene.Here,we introduce PlantGaussian,an approach for generating realistic three-dimensional(3D)visualization for plants across time and scenes.It marks one of the first applications of 3D Gaussian splatting techniques in plant science,achieving high-quality visualization across species and growth stages.By integrating the Segment Anything Model(SAM)and tracking algorithms,PlantGaussian overcomes the limitations of classic Gaussian reconstruction techniques in complex planting environments.A new mesh partitioning technique is employed to convert Gaussian rendering results into measurable plant meshes,offering a methodology for accurate 3D plant morphology phenotyping.To support this approach,PlantGaussian dataset is developed,which includes images of four crop species captured under multiple conditions and growth stages.Using only plant image sequences as input,it computes high-fidelity plant visualization models and 3D meshes for 3D plant morphological phenotyping.Visualization results indicate that most plant models achieve a Peak Signal-to-Noise Ratio(PSNR)exceeding 25,outperforming all models including the original 3D Gaussian Splatting and enhanced NeRF.The mesh results indicate an average relative error of 4%between the calculated values and the true measurements.As a generic 3D digital plant model,PlantGaussian will support expansion of plant phenotype databases,ecological research,and remote expert consultations.
基金supported by the Central Program of Basic Science of the National Natural Science Foundation of China(No.72088101)"The theory and application of resource and environment management in the digital economy era"+1 种基金The National Natural Science Foundation of China(No.41941018)Scientific research and technological development program of RIPED,"major research of basic geologic and synergy research of engineering practice on Gulong shale oil"(No.2021ycq01).
文摘Reliable prediction of the shale fracturing process is a challenging problem in exploiting deep shale oil and gas resources.Complex fracture networks need to be artificially created to employ deep shale oil and gas reserves.Randomly distributed minerals and heterogeneities in shales significantly affect mechanical properties and fracturing behaviors in oil and gas exploitation.Describing the actual microstructure and associated heterogeneities in shales constitutes a significant challenge.The RFPA3D(rock failure process analysis parallel computing program)-based modeling approach is a promising numerical technique due to its unique capability to simulate the fracturing behavior of rocks.To improve traditional numerical technology and study crack propagation in shale on the microscopic scale,a combination of high-precision internal structure detection technology with the RFPA^(3D) numerical simulation method was developed to construct a real mineral structure-based modeling method.First,an improved digital image processing technique was developed to incorporate actual shale microstructures(focused ion beam scanning electron microscopy was used to capture shale microstructure images that reflect the distri-butions of different minerals)into the numerical model.Second,the effect of mineral inhomogeneity was considered by integrating the mineral statistical model obtained from the mineral nanoindentation experiments into the numerical model.By simulating a shale numerical model in which pyrite particles are wrapped by organic matter,the effects of shale microstructure and applied stress state on microcrack behavior and mechanical properties were investigated and analyzed.In this study,the effect of pyrite particles on fracture propagation was systematically analyzed and summarized for the first time.The results indicate that the distribution of minerals and initial defects dominated the fracture evolution and the failure mode.Cracks are generally initiated and propagated along the boundaries of hard mineral particles such as pyrite or in soft minerals such as organic matter.Locations with collections of hard minerals are more likely to produce complex fractures.This study provides a valuable method for un-derstanding the microfracture behavior of shales.
文摘It is a key feature to embed 3D realistic sound effect in the future multimedia and virtual reality systems. Recent research on acoustics and psychoacoustics reveals the important cues for sound localization and sound perception. One promising approach to generate 3D realistic sound effect uses two earphones by simulating the sound waveforms from sound source to eardrum. This paper summarizes two methods for generating 3D realistic sound and points out their inherent drawbacks. To overcome these drawbacks we propose a simplified model to generate 3D realistic sound at any positions in the horizontal plane based on the results of sound perception and localization. Experimental results show that the model is correct and efficient.
