In this paper, an approach to predicting randomly-shaped particle volume based on its two- Dimensional （2-D） digital image is explored. Conversion of gray-scale image of the particles to its binary counterpart is fi...In this paper, an approach to predicting randomly-shaped particle volume based on its two- Dimensional （2-D） digital image is explored. Conversion of gray-scale image of the particles to its binary counterpart is first performed using backlighting technique. The silhouette of particle is thus obtained, and consequently, informative features such as particle area, centroid and shape-related descriptors are collected. Several dimensionless parameters are defined, and used as regressor variables in a multiple linear regression model to predict particle volume. Regressor coefficients are found by fitting to a randomly selected sample of 501 panicles ranging in size from 4.75mm to 25ram. The model testing experiment is conducted against a different aggregate sample of the similar statistical properties, the errors of the model-predicted volume of the batch is within ±2%.展开更多
The volumetric variability of dry tropical forests in Brazil and the scarcity of studies on the subject show the need for the development of techniques that make it possible to obtain adequate and accurate wood volume...The volumetric variability of dry tropical forests in Brazil and the scarcity of studies on the subject show the need for the development of techniques that make it possible to obtain adequate and accurate wood volume estimates.In this study,we analyzed a database of thinning trees from a forest management plan in the Contendas de SincoráNational Forest,southwestern Bahia State,Brazil.The data set included a total of 300 trees with a trunk diameter ranging from 5 to 52 cm.Adjustments,validation and statistical selection of four volumetric models were performed.Due to the difference in height values for the same diameter and the low correlation between both variables,we do not suggest models which only use the diameter at breast height(DBH)variable as a predictor because they accommodate the largest estimation errors.In comparing the best single entry model(Hohenald-Krenn)with the Spurr model(best fit model),it is noted that the exclusion of height as a predictor causes the values of 136.44 and 0.93 for Akaike information criterion(AIC)and adjusted determination coefficient(R2 adj),which are poorer than the second best model(Schumacher-Hall).Regarding the minimum sample size,errors in estimation(root mean square error(RMSE)and bias)of the best model decrease as the sample size increases,especially when a larger number of trees with DBH≥15.0 cm are randomly sampled.Stratified sampling by diameter class produces smaller volume prediction errors than random sampling,especially when considering all trees.In summary,the Spurr and Schumacher-Hall models perform better.These models suggest that the total variance explained in the estimates is not less than 95%,producing reliable forecasts of the total volume with shell.Our estimates indicate that the bias around the average is not greater than 7%.Our results support the decision to use regression methods to build models and estimate their parameters,seeking stratification strategies in diameter classes for the sample trees.Volume estimates with valid confidence intervals can be obtained using the Spurr model for the studied dry forest.Stratified sampling of the data set for model adjustment and selection is necessary,since we find significant results with mean error square root values and bias of up to 70%of the total database.展开更多
Hybrid approaches such as combining video data with pure physics-based simulation have been popular in the recent decade for computer graphics. The key motivation is to clearly retain salient advantages from both data...Hybrid approaches such as combining video data with pure physics-based simulation have been popular in the recent decade for computer graphics. The key motivation is to clearly retain salient advantages from both data-driven method and model-centric numerical simulation, while overcoming certain difficulties of both. The Eulerian method, which has been widely employed in flow simulation, stores variables such as velocity and density on regular Cartesian grids, thereby it could be associated with （volumetric） video data on the same domain. This paper proposes a novel method for flow simulation, which is tightly coupling video-based reconstruction with physically-based simulation and making use of meaningful physical attributes during re-simulation. First, we reconstruct the density field from a single-view video. Second, we estimate the velocity field using the reconstructed density field as prior. In the iterative process, the pressure projection can be treated as a physical constraint and the results of each step are corrected by obtained velocity field in the Eulerian framework. Third, we use the reconstructed density field and velocity field to guide the Eulerian simulation with anticipated new results. Through the guidance of video data, we can produce new flows that closely match with the real scene exhibited in data acquisition. Moreover, in the multigrid Eulerian simulation, we can generate new visual effects which cannot be created from raw video acquisition, with a goal of easily producing many more visually interesting results and respecting true physical attributes at the same time. We demonstrate salient advantages of our hybrid method with a variety of animation examples.展开更多
基金Funded by the Zhejiang Provincial Educatrion Ministry (No.2004884), and the Scientific Research Start-up Foundation of Ningbo University (No.2004037).
文摘In this paper, an approach to predicting randomly-shaped particle volume based on its two- Dimensional （2-D） digital image is explored. Conversion of gray-scale image of the particles to its binary counterpart is first performed using backlighting technique. The silhouette of particle is thus obtained, and consequently, informative features such as particle area, centroid and shape-related descriptors are collected. Several dimensionless parameters are defined, and used as regressor variables in a multiple linear regression model to predict particle volume. Regressor coefficients are found by fitting to a randomly selected sample of 501 panicles ranging in size from 4.75mm to 25ram. The model testing experiment is conducted against a different aggregate sample of the similar statistical properties, the errors of the model-predicted volume of the batch is within ±2%.
基金the National Council for Scientific and Technological Development-CNPq for granting financial support to the project(484260/2013-8).
文摘The volumetric variability of dry tropical forests in Brazil and the scarcity of studies on the subject show the need for the development of techniques that make it possible to obtain adequate and accurate wood volume estimates.In this study,we analyzed a database of thinning trees from a forest management plan in the Contendas de SincoráNational Forest,southwestern Bahia State,Brazil.The data set included a total of 300 trees with a trunk diameter ranging from 5 to 52 cm.Adjustments,validation and statistical selection of four volumetric models were performed.Due to the difference in height values for the same diameter and the low correlation between both variables,we do not suggest models which only use the diameter at breast height(DBH)variable as a predictor because they accommodate the largest estimation errors.In comparing the best single entry model(Hohenald-Krenn)with the Spurr model(best fit model),it is noted that the exclusion of height as a predictor causes the values of 136.44 and 0.93 for Akaike information criterion(AIC)and adjusted determination coefficient(R2 adj),which are poorer than the second best model(Schumacher-Hall).Regarding the minimum sample size,errors in estimation(root mean square error(RMSE)and bias)of the best model decrease as the sample size increases,especially when a larger number of trees with DBH≥15.0 cm are randomly sampled.Stratified sampling by diameter class produces smaller volume prediction errors than random sampling,especially when considering all trees.In summary,the Spurr and Schumacher-Hall models perform better.These models suggest that the total variance explained in the estimates is not less than 95%,producing reliable forecasts of the total volume with shell.Our estimates indicate that the bias around the average is not greater than 7%.Our results support the decision to use regression methods to build models and estimate their parameters,seeking stratification strategies in diameter classes for the sample trees.Volume estimates with valid confidence intervals can be obtained using the Spurr model for the studied dry forest.Stratified sampling of the data set for model adjustment and selection is necessary,since we find significant results with mean error square root values and bias of up to 70%of the total database.
基金This work was supported by the National Natural Science Foundation of China under Grant Nos. 61532002, 61672237, 61672077 and 61672149, the Natural Science Foundation of USA under Grant Nos. IIS-1715985, IIS-0949467, IIS-1047715, and IIS-1049448, and the National High Technology Research and Development 863 Program of China under Grant No. 2015AA016404.
文摘Hybrid approaches such as combining video data with pure physics-based simulation have been popular in the recent decade for computer graphics. The key motivation is to clearly retain salient advantages from both data-driven method and model-centric numerical simulation, while overcoming certain difficulties of both. The Eulerian method, which has been widely employed in flow simulation, stores variables such as velocity and density on regular Cartesian grids, thereby it could be associated with （volumetric） video data on the same domain. This paper proposes a novel method for flow simulation, which is tightly coupling video-based reconstruction with physically-based simulation and making use of meaningful physical attributes during re-simulation. First, we reconstruct the density field from a single-view video. Second, we estimate the velocity field using the reconstructed density field as prior. In the iterative process, the pressure projection can be treated as a physical constraint and the results of each step are corrected by obtained velocity field in the Eulerian framework. Third, we use the reconstructed density field and velocity field to guide the Eulerian simulation with anticipated new results. Through the guidance of video data, we can produce new flows that closely match with the real scene exhibited in data acquisition. Moreover, in the multigrid Eulerian simulation, we can generate new visual effects which cannot be created from raw video acquisition, with a goal of easily producing many more visually interesting results and respecting true physical attributes at the same time. We demonstrate salient advantages of our hybrid method with a variety of animation examples.
