Rendering, storing and transmission of high resolution models composed of large number of triangles primitives represents one of problems that scientific visualization and virtual reality faces. When these models exce...
详细信息
Edge detection is an important problem in image processing. This was extensively studied in recent years. In this paper artificial feedforward neural networks are used for identifying edges in gray-scale images. Super...
详细信息
The main objective of this work is the design and implementation of a hierarchical control solution for chemical and petrochemical processes, namely the control and optimization of the ethylene reactor, the key-instal...
详细信息
The main objective of this work is the design and implementation of a hierarchical control solution for chemical and petrochemical processes, namely the control and optimization of the ethylene reactor, the key-installation in the petrochemical industry. The hierarchical control structures are organized on two levels of automation. The execution or control level, on the one hand, has to fulfill two important tasks: the data acquisition from the physical plant and the regulation of the major parameters of the process. The supervisory level, on the other hand, is referred to the optimization of the production.
Computed Tomography (CT) scanners have a wide variety of uses, particularly in medical and industrial imaging. However, the data generated by these devices does not have intrinsic visual information and is often large...
详细信息
Computed Tomography (CT) scanners have a wide variety of uses, particularly in medical and industrial imaging. However, the data generated by these devices does not have intrinsic visual information and is often large enough to make it time consuming and tedious for an expert to manually inspect. We present techniques which allow for the fast and intuitive semi-automatic visual analysis of large CT data, comprising hundreds of slices. These techniques make use of the vast capabilities of modern GPUs, allowing for large bodies of data to be processed in real time, while generating images where the information obtained via CT scans is classified and represented through a precise control of color and transparency. We provide detailed descriptions of the underlying algorithms and methods involved in processing the CT information, present our results via images generated by our proprietary rendering frameworks, and explain how these images convey a faster and more intuitive means of inspecting the underlying data.
Edge detection is an important problem in image processing. This was extensively studied in recent years. In this paper artificial feedforward neural networks are used for identifying edges in gray-scale images. Super...
详细信息
Edge detection is an important problem in image processing. This was extensively studied in recent years. In this paper artificial feedforward neural networks are used for identifying edges in gray-scale images. Supervised learning based on the gradient descent algorithm is used. A new method is proposed for neural network training patterns using fuzzy concepts. Fuzzy membership functions are used for improving the generalization capability of neural network. Edge detection in noisy images without applying a noise removal technique poses a difficult problem. The proposed method obtained successful experimental results for digital images. The advantage of our approach is that we use the trained neural network as a filter on both noisy and noise free images.
Rendering, storing and transmission of high resolution models composed of large number of triangles primitives represents one of problems that scientific visualization and virtual reality faces. When these models exce...
详细信息
Rendering, storing and transmission of high resolution models composed of large number of triangles primitives represents one of problems that scientific visualization and virtual reality faces. When these models exceed the capacity of the graphics hardware real-time rendering the solution is to simplify the mesh models by eliminating primitives from the original models while trying to keep topological information and main surface characteristics. This paper presents a new approach to the quadric error metric mesh simplification that uses the local surface features to quantify the cost of an edge collapse. We achieve this by using a weight for each quadric error processed that takes into account local Willmore energy. We defined the probability of each vertex to be decimated or not in the simplification process to be the entropy value of the vertex neighborhood Willmore energy values. We demonstrate that our method can obtain better geometric results than QSlim simplifications by preserving local features in regions of high importance for the viewer.
In the context of wireless networked control systems energy constraints may become crucial.A special way to cope with them is based on an additional degree of freedom by independently selecting the sampling period and...
详细信息
ISBN:
(纸本)9781467325813
In the context of wireless networked control systems energy constraints may become crucial.A special way to cope with them is based on an additional degree of freedom by independently selecting the sampling period and the transmission *** transmission intervals are introduced to reduce the communication costs while preserving the quality of *** this paper a model-free and a novel model-based structure towards integral action are introduced.A stability proof for both approaches is given.
Due to the recent research advances on quantum computing, ideas from this field have been increasingly used as a source of inspiration for new variants of evolutionary algorithms. In this paper, the QIEA-SSEHC algorit...
详细信息
Due to the recent research advances on quantum computing, ideas from this field have been increasingly used as a source of inspiration for new variants of evolutionary algorithms. In this paper, the QIEA-SSEHC algorithm is proposed for solving multi-attribute combinatorial auction problems in multi-agent systems, characterized by an evolutionary hill-climbing phase, a steady state model and a repair procedure to keep all the individuals feasible. The results are compared to those of NSGA-II, a well-known multi-objective evolutionary algorithm, and convergence and diversity metrics are used to assess the quality of multidimensional solutions.
Usually when modeling a real process the effects of nonlinearities must be taken into account. A road traffic sector is one such process. Considering it from a macroscopic point of view, the sector is modeled both as ...
详细信息
Usually when modeling a real process the effects of nonlinearities must be taken into account. A road traffic sector is one such process. Considering it from a macroscopic point of view, the sector is modeled both as a nonlinear and as a linearized process. An 110 model is proposed and a RST control algorithm is used to ensure the imposed performances for the closed loop system. The robustness of the system is obtained by determining the modulus margin and based upon it the admissible nonlinearities are analyzed in order to determine the limits between which the system maintains its stability.
A difficult challenge in hybrid electric vehicles (HEVs) and full electric vehicles (EVs) is the torque control of externally excited synchronous machines (EESMs). Effective torque control requires an efficient soluti...
详细信息
A difficult challenge in hybrid electric vehicles (HEVs) and full electric vehicles (EVs) is the torque control of externally excited synchronous machines (EESMs). Effective torque control requires an efficient solution to the state reference generation problem, which is a nonlinear non-convex optimization problem. The goal of this paper is to develop a state reference generation algorithm based on the griding of the state and output spaces. Firstly, an approximation defined over a cubic partition of the torque function with a piecewise affine (PWA) function is made. As a result, the state reference generation problem is reduced in each cube to solving a convex optimization problem. Moreover, this approach provides guarantees about the error bound introduced by the state reference generation procedure for the full operational state-space. To illustrate the effectiveness and robustness of the proposed algorithm several simulation results are presented. The obtained results show significant improvement compared with existing state-of-the-art reference generation methods.
暂无评论