The Experts below are selected from a list of 321525 Experts worldwide ranked by ideXlab platform
Longin Jan Latecki - One of the best experts on this subject based on the ideXlab platform.
-
shape matching and classification using height functions
Pattern Recognition Letters, 2012Co-Authors: Junwei Wang, Longin Jan LateckiAbstract:We propose a novel shape descriptor for matching and recognizing 2D object silhouettes. The contour of each object is represented by a fixed number of Sample Points. For each Sample Point, a height function is defined based on the distances of the other Sample Points to its tangent line. One compact and robust shape descriptor is obtained by smoothing the height functions. The proposed descriptor is not only invariant to geometric transformations such as translation, rotation and scaling but also insensitive to nonlinear deformations due to noise and occlusion. In the matching stage, the Dynamic Programming (DP) algorithm is employed to find out the optimal correspondence between Sample Points of every two shapes. The height function provides an excellent discriminative power, which is demonstrated by excellent retrieval performances on several popular shape benchmarks, including MPEG-7 data set, Kimia's data set and ETH-80 data set.
-
shape matching and classification using height functions
Pattern Recognition Letters, 2012Co-Authors: Junwei Wang, Xiang Bai, Xinge You, Wenyu Liu, Longin Jan LateckiAbstract:We propose a novel shape descriptor for matching and recognizing 2D object silhouettes. The contour of each object is represented by a fixed number of Sample Points. For each Sample Point, a height function is defined based on the distances of the other Sample Points to its tangent line. One compact and robust shape descriptor is obtained by smoothing the height functions. The proposed descriptor is not only invariant to geometric transformations such as translation, rotation and scaling but also insensitive to nonlinear deformations due to noise and occlusion. In the matching stage, the Dynamic Programming (DP) algorithm is employed to find out the optimal correspondence between Sample Points of every two shapes. The height function provides an excellent discriminative power, which is demonstrated by excellent retrieval performances on several popular shape benchmarks, including MPEG-7 data set, Kimia's data set and ETH-80 data set.
Junwei Wang - One of the best experts on this subject based on the ideXlab platform.
-
shape matching and classification using height functions
Pattern Recognition Letters, 2012Co-Authors: Junwei Wang, Longin Jan LateckiAbstract:We propose a novel shape descriptor for matching and recognizing 2D object silhouettes. The contour of each object is represented by a fixed number of Sample Points. For each Sample Point, a height function is defined based on the distances of the other Sample Points to its tangent line. One compact and robust shape descriptor is obtained by smoothing the height functions. The proposed descriptor is not only invariant to geometric transformations such as translation, rotation and scaling but also insensitive to nonlinear deformations due to noise and occlusion. In the matching stage, the Dynamic Programming (DP) algorithm is employed to find out the optimal correspondence between Sample Points of every two shapes. The height function provides an excellent discriminative power, which is demonstrated by excellent retrieval performances on several popular shape benchmarks, including MPEG-7 data set, Kimia's data set and ETH-80 data set.
-
shape matching and classification using height functions
Pattern Recognition Letters, 2012Co-Authors: Junwei Wang, Xiang Bai, Xinge You, Wenyu Liu, Longin Jan LateckiAbstract:We propose a novel shape descriptor for matching and recognizing 2D object silhouettes. The contour of each object is represented by a fixed number of Sample Points. For each Sample Point, a height function is defined based on the distances of the other Sample Points to its tangent line. One compact and robust shape descriptor is obtained by smoothing the height functions. The proposed descriptor is not only invariant to geometric transformations such as translation, rotation and scaling but also insensitive to nonlinear deformations due to noise and occlusion. In the matching stage, the Dynamic Programming (DP) algorithm is employed to find out the optimal correspondence between Sample Points of every two shapes. The height function provides an excellent discriminative power, which is demonstrated by excellent retrieval performances on several popular shape benchmarks, including MPEG-7 data set, Kimia's data set and ETH-80 data set.
Antonio Fernando Branco Costa - One of the best experts on this subject based on the ideXlab platform.
-
a shewhart chart with alternated charting statistic to control multivariate poisson processes
Computers & Industrial Engineering, 2020Co-Authors: Roberto Campos Leoni, Antonio Fernando Branco CostaAbstract:Abstract In this article, a Shewhart c chart with alternating charting statistic (ACS Poisson chart) is proposed to control multivariate Poisson processes. Considering the bivariate case, where we have the occurrence of X and Y defects; if the current Sample Point is cx (cy), the number of X (Y) defects, then the next Sample Point will be cy (cx), the number of Y (X) defects. The ACS Poisson chart outperforms all its competitors, that is, the joint cx and cy Poisson charts and the bivariate Poisson charts with the following monitoring statistics: X + Y, X-Y, and the maximum between X and Y. At each sampling Point, the ACS Poisson chart controls the occurrences of only one type of non-conformity, because of that its unit of inspection might be twice larger. The ACS Poisson chart requires larger units of inspection to be more sensitive than its competitors, but not necessarily twice larger. The CUSUM version of the ACS Poisson chart also proved to be more efficient than the bivariate Poisson CUSUM chart. Similar results were also observed with the monitoring of trivariate Poisson processes.
-
the shewhart attribute chart with alternated charting statistics to monitor bivariate and trivariate mean vectors
Computers & Industrial Engineering, 2018Co-Authors: Roberto Campos Leoni, Antonio Fernando Branco CostaAbstract:Abstract In this article, we combined the Alternated Charting Statistic (ACS) scheme with the traditional attribute np chart to control mean vectors of bivariate and trivariate normal processes. With the bivariate ACS scheme in use (the trivariate scheme is similar), the two quality characteristics (X, Y) are controlled in an alternating fashion. If the current Sample Point is the number of disapproved items with respect to the X discriminating limits, then the next Sample Point will be the number of disapproved items with respect to the Y discriminating limits. The strategy of using the X discriminating limits to classify the items of one Sample and the Y discriminating limits to classify the items of the next Sample instead of using jointly the X and Y discriminating limits to classify the items of all Samples might be compensated with the adoption of larger Samples. In other words, the proposed bivariate (trivariate) ACS chart might work with Samples as large as 2n (3n); n is the Sample size of the competing Hotelling and Max D charts. The proposed chart resembles an np chart with alternated charting statistic; because of that, it is called the ACS mp chart. The ACS mp chart always outperforms the Max D chart and, in comparison with the standard T2 chart and with the combined Max D − T2 chart, it has a better overall performance. With the ACS scheme, the items are classified as approved or disapproved regarding only one of the two quality characteristic, X or Y; with the Max D chart the complexity increases, once the items are classified into four different categories: approved (disapproved) regarding both, the X and Y discriminating limits, or approved (disapproved) regarding the X discriminate limits and disapproved (approved) regarding the Y discriminate limits. The T2 chart always requires the measurement of the two quality characteristics. The additional advantage of inspecting only one quality characteristic of the Sample items lies in the fact that the XY-correlation doesn’t need to be estimated.
-
the s chart with variable charting statistic to control bi and trivariate processes
Computers & Industrial Engineering, 2017Co-Authors: Antonio Fernando Branco Costa, Antonio Faria NetoAbstract:Abstract In this article, we propose the S chart with variable charting statistic to control the covariance matrix as an alternative to the use of the bivariate |S| chart and the trivariate VMAX chart. As usual, Samples are regularly taken from the process, but only one of the two quality characteristics, X or Y, is measured and only one of the two statistics ( S x , S y ) is computed. The statistic in use and the position of the current Sample Point on the chart define the statistic for the next Sample. If the current Point is the standard deviation of the X values and it is in the central region (warning region), then the statistic for the next Sample will be the standard deviation of the Y values (X values). If the current Point is the standard deviation of the Y values and it is in the central region (warning region), then the statistic for the next Sample will be the standard deviation of the X values (Y values). For the trivariate case, when the Sample Point falls in the central region, the charting statistic for the next Sample changes from S x to S y , or from S y to S z , or yet, from S z to S x . The VCS chart is not only operationally simpler than the bivariate |S| and trivariate VMAX charts but also signals faster even with less measurements per Sample.
Meunier Gérard - One of the best experts on this subject based on the ideXlab platform.
-
Adaptive MultiPoint Model Order Reduction Scheme for Large- Scale Inductive PEEC Circuits
'Institute of Electrical and Electronics Engineers (IEEE)', 2017Co-Authors: Nguyen Trung-son, Tung Le Duc, Tran Thanh-son, Guichon Jean-michel, Chadebec Olivier, Meunier GérardAbstract:WOS:000398974400015International audienceThe model order reduction techniques based on the multiPoint projection Krylov methods have become the methods of choice to generate large macromodels of multiport RLC circuits. A well-known difficulty with such methods lies in the need for clever Point selection to attain model compactness and accuracy. In this paper, we present an automatic methodology for optimizing Sample Point selection. This method is general, suitable for circuits obtained by the partial element equivalent circuit method coupled with an adaptive multilevel fast multipole method. Our algorithm has been validated on an industrial example to demonstrate the accuracy and robustness of the approach
Gérard Meunier - One of the best experts on this subject based on the ideXlab platform.
-
Adaptive MultiPoint Model Order Reduction Scheme for Large-Scale Inductive PEEC Circuits
IEEE Transactions on Electromagnetic Compatibility, 2017Co-Authors: Trung-son Nguyen, Thanh-son Tran, Olivier Chadebec, Jean-michel Guichon, Gérard MeunierAbstract:The model order reduction techniques based on the multiPoint projection Krylov methods have become the methods of choice to generate large macromodels of multiport RLC circuits. A well-known difficulty with such methods lies in the need for clever Point selection to attain model compactness and accuracy. In this paper, we present an automatic methodology for optimizing Sample Point selection. This method is general, suitable for circuits obtained by the partial element equivalent circuit method coupled with an adaptive multilevel fast multipole method. Our algorithm has been validated on an industrial example to demonstrate the accuracy and robustness of the approach.
