The appearance of a face is severely altered by illumination conditions that makes automatic face recognition a challenging task. In this paper we propose a Gaussian Mixture Models (GMM)-based human face identificatio...The appearance of a face is severely altered by illumination conditions that makes automatic face recognition a challenging task. In this paper we propose a Gaussian Mixture Models (GMM)-based human face identification technique built in the Fourier or frequency domain that is robust to illumination changes and does not require “illumination normalization” (removal of illumination effects) prior to application unlike many existing methods. The importance of the Fourier domain phase in human face identification is a well-established fact in signal processing. A maximum a posteriori (or, MAP) estimate based on the posterior likelihood is used to perform identification, achieving misclassification error rates as low as 2% on a database that contains images of 65 individuals under 21 different illumination conditions. Furthermore, a misclassification rate of 3.5% is observed on the Yale database with 10 people and 64 different illumination conditions. Both these sets of results are significantly better than those obtained from traditional PCA and LDA classifiers. Statistical analysis pertaining to model selection is also presented.展开更多
Representation of roughness is introduced and the rationality of applying thephysics-based model in RE is analyzed at first. Then the scattering theory of electromagnetic waveis simplified and deduced to a physics-bas...Representation of roughness is introduced and the rationality of applying thephysics-based model in RE is analyzed at first. Then the scattering theory of electromagnetic waveis simplified and deduced to a physics-based model according to the characteristics of the surfaceto be reconstructed in RE. At last, the intensity diagrams of reflected field distribution areprovided to prove the feasibility of the presented model and some spheres are rendered with thismodel.展开更多
With the coming of information age and the development of computer science, digitalization of whole garment is becoming more and more important. The surface of whole garment is sequent and glossy so that it is lack of...With the coming of information age and the development of computer science, digitalization of whole garment is becoming more and more important. The surface of whole garment is sequent and glossy so that it is lack of the texture characteristic which is the key of digital 3D modeling. According to this reason, the structure illumination is steered into a method of this paper. The paper proposes the method by which 3D model of whole garment is created from 2D image sequences directly but not by the common techniques using general CAD model. In the paper the structure illumination is generated by the slide projector and the modeling of whole garment is based on the strict theory of the digital photogrammetry, computer vision and image processing pattern recognition. Because whole garment is lack of the applicable texture for matching, the characteristic texture generated by the structure illumination is added onto the surface of whole garment. After the characteristic texture is extracted from images and is matched well, 3D coordinates of the characteristic texture can be calculated out by the space forward intersection. Then the whole garment model is acquired by connecting all neighbour space points in the TIN and rendering the real texture of whole garment automatically. The 3D modeling method is untouched so that it is nondestructive which is just suitable for the messaline and the clothing. The method of whole garment 3D modeling proposed in the paper is flexible, effective and practical, which is confirmed by the results of the reconstructing experiments.展开更多
Optimum laser configurations are presented to achieve high illumination uniformity with directly driven inertial confinement fusion targets.Assuming axisymmetric absorption pattern of individual laser beams,theoretica...Optimum laser configurations are presented to achieve high illumination uniformity with directly driven inertial confinement fusion targets.Assuming axisymmetric absorption pattern of individual laser beams,theoretical models are reviewed in terms of the number of laser beams,system imperfection,and laser beam patterns.Utilizing a self-organizing system of charged particles on a sphere,a simple numerical model is provided to give an optimal configuration for an arbitrary number of laser beams.As a result,such new configurations as“M48”and“M60”are found to show substantially higher illumination uniformity than any other existing direct drive systems.A new polar direct-drive scheme is proposed with the laser axes keeping off the target center,which can be applied to laser configurations designed for indirectly driven inertial fusion.展开更多
文摘The appearance of a face is severely altered by illumination conditions that makes automatic face recognition a challenging task. In this paper we propose a Gaussian Mixture Models (GMM)-based human face identification technique built in the Fourier or frequency domain that is robust to illumination changes and does not require “illumination normalization” (removal of illumination effects) prior to application unlike many existing methods. The importance of the Fourier domain phase in human face identification is a well-established fact in signal processing. A maximum a posteriori (or, MAP) estimate based on the posterior likelihood is used to perform identification, achieving misclassification error rates as low as 2% on a database that contains images of 65 individuals under 21 different illumination conditions. Furthermore, a misclassification rate of 3.5% is observed on the Yale database with 10 people and 64 different illumination conditions. Both these sets of results are significantly better than those obtained from traditional PCA and LDA classifiers. Statistical analysis pertaining to model selection is also presented.
文摘Representation of roughness is introduced and the rationality of applying thephysics-based model in RE is analyzed at first. Then the scattering theory of electromagnetic waveis simplified and deduced to a physics-based model according to the characteristics of the surfaceto be reconstructed in RE. At last, the intensity diagrams of reflected field distribution areprovided to prove the feasibility of the presented model and some spheres are rendered with thismodel.
文摘With the coming of information age and the development of computer science, digitalization of whole garment is becoming more and more important. The surface of whole garment is sequent and glossy so that it is lack of the texture characteristic which is the key of digital 3D modeling. According to this reason, the structure illumination is steered into a method of this paper. The paper proposes the method by which 3D model of whole garment is created from 2D image sequences directly but not by the common techniques using general CAD model. In the paper the structure illumination is generated by the slide projector and the modeling of whole garment is based on the strict theory of the digital photogrammetry, computer vision and image processing pattern recognition. Because whole garment is lack of the applicable texture for matching, the characteristic texture generated by the structure illumination is added onto the surface of whole garment. After the characteristic texture is extracted from images and is matched well, 3D coordinates of the characteristic texture can be calculated out by the space forward intersection. Then the whole garment model is acquired by connecting all neighbour space points in the TIN and rendering the real texture of whole garment automatically. The 3D modeling method is untouched so that it is nondestructive which is just suitable for the messaline and the clothing. The method of whole garment 3D modeling proposed in the paper is flexible, effective and practical, which is confirmed by the results of the reconstructing experiments.
基金This work was supported by the Japan Society for the Promotion of Science(JSPS).
文摘Optimum laser configurations are presented to achieve high illumination uniformity with directly driven inertial confinement fusion targets.Assuming axisymmetric absorption pattern of individual laser beams,theoretical models are reviewed in terms of the number of laser beams,system imperfection,and laser beam patterns.Utilizing a self-organizing system of charged particles on a sphere,a simple numerical model is provided to give an optimal configuration for an arbitrary number of laser beams.As a result,such new configurations as“M48”and“M60”are found to show substantially higher illumination uniformity than any other existing direct drive systems.A new polar direct-drive scheme is proposed with the laser axes keeping off the target center,which can be applied to laser configurations designed for indirectly driven inertial fusion.
