The Experts below are selected from a list of 153 Experts worldwide ranked by ideXlab platform
Kacem Chehdi - One of the best experts on this subject based on the ideXlab platform.
-
efficient rotation scaling Translation Parameter estimation based on the fractal image model
IEEE Transactions on Geoscience and Remote Sensing, 2016Co-Authors: Benoit Vozel, Vladimir V Lukin, Kacem ChehdiAbstract:This paper deals with area-based subpixel image registration under the rotation-isometric scaling-Translation transformation hypothesis. Our approach is based on parametrical modeling of geometrically transformed textural image fragments and maximum-likelihood estimation of the transformation vector between them. Due to the parametrical approach based on the fractional Brownian motion modeling of the local fragments' texture, the proposed estimator $\mbox{ML}_\mathrm{fBm}$ (ML stands for “maximum likelihood” and fBm stands for “fractal Brownian motion”) has the ability to better adapt to real image texture content compared with other methods relying on universal similarity measures such as mutual information or normalized correlation. The main benefits are observed when assumptions underlying the fBm model are fully satisfied, e.g., for isotropic normally distributed textures with stationary increments. Experiments on both simulated and real images and for high and weak correlations between registered images show that the $\mbox{ML}_\mathrm{fBm}$ estimator offers significant improvement compared with other state-of-the-art methods. It reduces Translation vector, rotation angle, and scaling factor estimation errors by a factor of about 1.75–2, and it decreases the probability of false match by up to five times. In addition, an accurate confidence interval for $\mbox{ML}_\mathrm{fBm}$ estimates can be obtained from the Cramer–Rao lower bound on rotation-scaling-Translation Parameter estimation error. This bound depends on texture roughness, noise level in reference and template images, correlation between these images, and geometrical transformation Parameters.
-
Efficient Rotation-Scaling-Translation Parameter Estimation Based on the Fractal Image Model
IEEE Transactions on Geoscience and Remote Sensing, 2016Co-Authors: Benoit Vozel, Vladimir V Lukin, Kacem ChehdiAbstract:This paper deals with area-based subpixel image registration under the rotation-isometric scaling-Translation transformation hypothesis. Our approach is based on parametrical modeling of geometrically transformed textural image fragments and maximum-likelihood estimation of the transformation vector between them. Due to the parametrical approach based on the fractional Brownian motion modeling of the local fragments' texture, the proposed estimator MLfBm (ML stands for “maximum likelihood” and fBm stands for “fractal Brownian motion”) has the ability to better adapt to real image texture content compared with other methods relying on universal similarity measures such as mutual information or normalized correlation. The main benefits are observed when assumptions underlying the fBm model are fully satisfied, e.g., for isotropic normally distributed textures with stationary increments. Experiments on both simulated and real images and for high and weak correlations between registered images show that the MLfBm estimator offers significant improvement compared with other state-of-the-art methods. It reduces Translation vector, rotation angle, and scaling factor estimation errors by a factor of about 1.75-2, and it decreases the probability of false match by up to five times. In addition, an accurate confidence interval for MLfBm estimates can be obtained from the Cramér-Rao lower bound on rotation-scaling-Translation Parameter estimation error. This bound depends on texture roughness, noise level in reference and template images, correlation between these images, and geometrical transformation Parameters.
Benoit Vozel - One of the best experts on this subject based on the ideXlab platform.
-
efficient rotation scaling Translation Parameter estimation based on the fractal image model
IEEE Transactions on Geoscience and Remote Sensing, 2016Co-Authors: Benoit Vozel, Vladimir V Lukin, Kacem ChehdiAbstract:This paper deals with area-based subpixel image registration under the rotation-isometric scaling-Translation transformation hypothesis. Our approach is based on parametrical modeling of geometrically transformed textural image fragments and maximum-likelihood estimation of the transformation vector between them. Due to the parametrical approach based on the fractional Brownian motion modeling of the local fragments' texture, the proposed estimator $\mbox{ML}_\mathrm{fBm}$ (ML stands for “maximum likelihood” and fBm stands for “fractal Brownian motion”) has the ability to better adapt to real image texture content compared with other methods relying on universal similarity measures such as mutual information or normalized correlation. The main benefits are observed when assumptions underlying the fBm model are fully satisfied, e.g., for isotropic normally distributed textures with stationary increments. Experiments on both simulated and real images and for high and weak correlations between registered images show that the $\mbox{ML}_\mathrm{fBm}$ estimator offers significant improvement compared with other state-of-the-art methods. It reduces Translation vector, rotation angle, and scaling factor estimation errors by a factor of about 1.75–2, and it decreases the probability of false match by up to five times. In addition, an accurate confidence interval for $\mbox{ML}_\mathrm{fBm}$ estimates can be obtained from the Cramer–Rao lower bound on rotation-scaling-Translation Parameter estimation error. This bound depends on texture roughness, noise level in reference and template images, correlation between these images, and geometrical transformation Parameters.
-
Efficient Rotation-Scaling-Translation Parameter Estimation Based on the Fractal Image Model
IEEE Transactions on Geoscience and Remote Sensing, 2016Co-Authors: Benoit Vozel, Vladimir V Lukin, Kacem ChehdiAbstract:This paper deals with area-based subpixel image registration under the rotation-isometric scaling-Translation transformation hypothesis. Our approach is based on parametrical modeling of geometrically transformed textural image fragments and maximum-likelihood estimation of the transformation vector between them. Due to the parametrical approach based on the fractional Brownian motion modeling of the local fragments' texture, the proposed estimator MLfBm (ML stands for “maximum likelihood” and fBm stands for “fractal Brownian motion”) has the ability to better adapt to real image texture content compared with other methods relying on universal similarity measures such as mutual information or normalized correlation. The main benefits are observed when assumptions underlying the fBm model are fully satisfied, e.g., for isotropic normally distributed textures with stationary increments. Experiments on both simulated and real images and for high and weak correlations between registered images show that the MLfBm estimator offers significant improvement compared with other state-of-the-art methods. It reduces Translation vector, rotation angle, and scaling factor estimation errors by a factor of about 1.75-2, and it decreases the probability of false match by up to five times. In addition, an accurate confidence interval for MLfBm estimates can be obtained from the Cramér-Rao lower bound on rotation-scaling-Translation Parameter estimation error. This bound depends on texture roughness, noise level in reference and template images, correlation between these images, and geometrical transformation Parameters.
Vladimir V Lukin - One of the best experts on this subject based on the ideXlab platform.
-
efficient rotation scaling Translation Parameter estimation based on the fractal image model
IEEE Transactions on Geoscience and Remote Sensing, 2016Co-Authors: Benoit Vozel, Vladimir V Lukin, Kacem ChehdiAbstract:This paper deals with area-based subpixel image registration under the rotation-isometric scaling-Translation transformation hypothesis. Our approach is based on parametrical modeling of geometrically transformed textural image fragments and maximum-likelihood estimation of the transformation vector between them. Due to the parametrical approach based on the fractional Brownian motion modeling of the local fragments' texture, the proposed estimator $\mbox{ML}_\mathrm{fBm}$ (ML stands for “maximum likelihood” and fBm stands for “fractal Brownian motion”) has the ability to better adapt to real image texture content compared with other methods relying on universal similarity measures such as mutual information or normalized correlation. The main benefits are observed when assumptions underlying the fBm model are fully satisfied, e.g., for isotropic normally distributed textures with stationary increments. Experiments on both simulated and real images and for high and weak correlations between registered images show that the $\mbox{ML}_\mathrm{fBm}$ estimator offers significant improvement compared with other state-of-the-art methods. It reduces Translation vector, rotation angle, and scaling factor estimation errors by a factor of about 1.75–2, and it decreases the probability of false match by up to five times. In addition, an accurate confidence interval for $\mbox{ML}_\mathrm{fBm}$ estimates can be obtained from the Cramer–Rao lower bound on rotation-scaling-Translation Parameter estimation error. This bound depends on texture roughness, noise level in reference and template images, correlation between these images, and geometrical transformation Parameters.
-
Efficient Rotation-Scaling-Translation Parameter Estimation Based on the Fractal Image Model
IEEE Transactions on Geoscience and Remote Sensing, 2016Co-Authors: Benoit Vozel, Vladimir V Lukin, Kacem ChehdiAbstract:This paper deals with area-based subpixel image registration under the rotation-isometric scaling-Translation transformation hypothesis. Our approach is based on parametrical modeling of geometrically transformed textural image fragments and maximum-likelihood estimation of the transformation vector between them. Due to the parametrical approach based on the fractional Brownian motion modeling of the local fragments' texture, the proposed estimator MLfBm (ML stands for “maximum likelihood” and fBm stands for “fractal Brownian motion”) has the ability to better adapt to real image texture content compared with other methods relying on universal similarity measures such as mutual information or normalized correlation. The main benefits are observed when assumptions underlying the fBm model are fully satisfied, e.g., for isotropic normally distributed textures with stationary increments. Experiments on both simulated and real images and for high and weak correlations between registered images show that the MLfBm estimator offers significant improvement compared with other state-of-the-art methods. It reduces Translation vector, rotation angle, and scaling factor estimation errors by a factor of about 1.75-2, and it decreases the probability of false match by up to five times. In addition, an accurate confidence interval for MLfBm estimates can be obtained from the Cramér-Rao lower bound on rotation-scaling-Translation Parameter estimation error. This bound depends on texture roughness, noise level in reference and template images, correlation between these images, and geometrical transformation Parameters.
Sai Ho Ling - One of the best experts on this subject based on the ideXlab platform.
-
Genetic Algorithm-Based VariableTranslation Wavelet NeuralNetwork anditsApplication
2020Co-Authors: Sai Ho Ling, F. H. F. LeungAbstract:A variable Translation wavelet neuralnetwork (VTWNN)trained bygenetic algorithm ispresented inthis paper.In theproposed waveletneuralnetwork, the Translation Parameters arevariables depending on the network inputs. Thanks tothevariable Translation Parameter, thenetworkbecomesan adaptive one,providing better performance andincreased learning ability thanconventional wavelet neural networks. Genetic algorithm isapplied totrain theParameters oftheproposed wavelet neural networkAn application exampleon short-term dailyelectric load forecasting inHongKongispresented toshowthemerits of theproposed network
-
Hybrid PSO-based variable Translation wavelet neural network and its application to hypoglycemia detection system
Neural Computing and Applications, 2012Co-Authors: Sai Ho Ling, Hung T. NguyenAbstract:To provide the detection of hypoglycemic episodes in Type 1 diabetes mellitus, hypoglycemia detection system is developed by the use of variable Translation wavelet neural network (VTWNN) in this paper. A wavelet neural network with variable Translation Parameter is selected as a suitable classifier because of its excellent characteristics in capturing nonstationary signal analysis and nonlinear function modeling. Due to the variable Translation Parameters, the network becomes an adaptive network and provides better classification performance. An improved hybrid particle swarm optimization is used to train the Parameters of VTWNN. Using the proposed classifier, a sensitivity of 81.40 % and a specificity of 50.91 % were achieved. The comparison results also show that the proposed detection system performs well in terms of good sensitivity and acceptable specificity.
-
Genetic algorithm-based variable Translation wavelet neural network and its application
Proceedings. 2005 IEEE International Joint Conference on Neural Networks 2005., 2005Co-Authors: Sai Ho Ling, F. H. F. LeungAbstract:A variable Translation wavelet neural network (VTWNN) trained by genetic algorithm is presented in this paper. In the proposed wavelet neural network, the Translation Parameters are variables depending on the network inputs. Thanks to the variable Translation Parameter, the network becomes an adaptive one, providing better performance and increased learning ability than conventional wavelet neural networks. Genetic algorithm is applied to train the Parameters of the proposed wavelet neural network. An application example on short-term daily electric load forecasting in Hong Kong is presented to show the merits of the proposed network.
F. H. F. Leung - One of the best experts on this subject based on the ideXlab platform.
-
Genetic Algorithm-Based VariableTranslation Wavelet NeuralNetwork anditsApplication
2020Co-Authors: Sai Ho Ling, F. H. F. LeungAbstract:A variable Translation wavelet neuralnetwork (VTWNN)trained bygenetic algorithm ispresented inthis paper.In theproposed waveletneuralnetwork, the Translation Parameters arevariables depending on the network inputs. Thanks tothevariable Translation Parameter, thenetworkbecomesan adaptive one,providing better performance andincreased learning ability thanconventional wavelet neural networks. Genetic algorithm isapplied totrain theParameters oftheproposed wavelet neural networkAn application exampleon short-term dailyelectric load forecasting inHongKongispresented toshowthemerits of theproposed network
-
Genetic algorithm-based variable Translation wavelet neural network and its application
Proceedings. 2005 IEEE International Joint Conference on Neural Networks 2005., 2005Co-Authors: Sai Ho Ling, F. H. F. LeungAbstract:A variable Translation wavelet neural network (VTWNN) trained by genetic algorithm is presented in this paper. In the proposed wavelet neural network, the Translation Parameters are variables depending on the network inputs. Thanks to the variable Translation Parameter, the network becomes an adaptive one, providing better performance and increased learning ability than conventional wavelet neural networks. Genetic algorithm is applied to train the Parameters of the proposed wavelet neural network. An application example on short-term daily electric load forecasting in Hong Kong is presented to show the merits of the proposed network.
