Load balancing is a key issue in large-scale Object-Based Storage systems. Many data replication and migration algorithms have been proposed for load balancing in distributed systems. However, the two operations, data...
详细信息
Load balancing is a key issue in large-scale Object-Based Storage systems. Many data replication and migration algorithms have been proposed for load balancing in distributed systems. However, the two operations, data replication and migration, are studied separately. An adaptive load balancing algorithm is presented in this paper, which combines replication and migration in a uniform ***, this algorithm uses a hybrid load metric which reflects short-term and long-term load status. To solve the online problem with a changing workload, the algorithm employs an adaptive mechanism to keep track of the characteristics of workloads. The simulation results show that the algorithm with object replication and migration can markedly reduce the overall response time
This paper describes the scale invariant feature transform (SIFT) method for rapid preregistration of medical image. This technique originates from Lowe's method wherein preregistration is achieved by matching the...
详细信息
This paper describes the scale invariant feature transform (SIFT) method for rapid preregistration of medical image. This technique originates from Lowe's method wherein preregistration is achieved by matching the corresponding keypoints between two images. The computational complexity has been reduced when we applied SIFT preregistration method before refined registration due to its O(n) exponential calculations. The features of SIFT are highly distinctive and invariant to image scaling and rotation, and partially invariant to change in illumination and contrast, it is robust and repeatable for cursorily matching two images. We also altered the descriptor so our method can deal with multimodality preregistration
Mutual information (MI) is an effective criterion for multi-modal image registration. However the traditional MI function only includes intensity information of images and lacks sufficient spatial information to accur...
详细信息
Mutual information (MI) is an effective criterion for multi-modal image registration. However the traditional MI function only includes intensity information of images and lacks sufficient spatial information to accurately measure the degree of alignment of two images, and besides, it is apt to be influenced by intensity interpolation, therefore presents many local maxima which frequently lead to misregistration. Our paper proposes a criterion of adaptive combination of intensity and gradient field mutual information (ACMI). Unlike the intensity MI computed from two original images, the gradient field MI of two images is calculated from their gradient code maps (GCM) constructed by coding the gradient field information of corresponding original image. Because of their complementary properties, these two MI functions are combined to form ACMI by a nonlinear weight function which can be adaptively regulated according to their performances and make the better dominant in the combination. Experimental results demonstrate that the ACMI outperforms the traditional MI and furthermore the former is much less sensitive than the latter to the reduction of resolution or overlapped region of images
This paper presents the multi-agent architecture for artificial transportation system. In this architecture, Petri net is used as basic model to represent agents. At an intersection, the agents are divided into two gr...
详细信息
This paper presents the multi-agent architecture for artificial transportation system. In this architecture, Petri net is used as basic model to represent agents. At an intersection, the agents are divided into two groups: one for the traffic-signal and the other for the vehicle flow, which are integrated to represent the behavior of this intersection. In addition, those agents can be used as the modularity to represent urban network of more scale. To coordinate different intersection agents, game theory is used to design coordination strategy between agents. The iterated elimination of strictly dominated strategies algorithm is presented to find Nash equilibrium
Conventional freehand 3D ultrasound imaging techniques separate the scanning and reconstruction steps from the visualization step. Several approaches that enable real-time volume reconstruction and visualization durin...
详细信息
Conventional freehand 3D ultrasound imaging techniques separate the scanning and reconstruction steps from the visualization step. Several approaches that enable real-time volume reconstruction and visualization during acquisition have been described in the literature in recent years. However, these approaches are not fully interactive, since they do not provide sufficient feedbacks on volume reconstruction. In this paper, we present a qualitative and quantitative interaction technique (QAQIT) which can provide accurate feedbacks on volume reconstruction in real time during acquisition. Two threads are designed for data acquisition: one is to acquire B-scan images from the ultrasound scanner sequentially, and the other is to obtain positions and orientations from the position sensor continuously. A matching thread matches the latest acquired image with its relative position and orientation (PAO), and then reconstructs the image into a predefined volume. After the 3D reconstruction of each image, we calculate the reconstructed ratio (RR) and increased ratio (IR) of the reconstruction volume (RV), then update the display of the RR, and drive the volume rendering of the RV according to the IR. A freehand system based on the QAQIT has been developed on a personal computer (PC). We demonstrate our system on an embryo phantom
This paper is concerned with reduced-order H∞ filtering of stochastic systems. Based on linear matrix inequality (LMI) technique, a new design method is proposed for the reduced-order filtering of stochastic linear s...
详细信息
In this paper, a new method based on relative difference space (RDS) and support vector machine (SVM) is proposed for multi-class recognition. First the RDS transformation converts the multi-class problem to a binary-...
详细信息
In this paper, a new method based on relative difference space (RDS) and support vector machine (SVM) is proposed for multi-class recognition. First the RDS transformation converts the multi-class problem to a binary-class problem, and then SVM is used for the binary classification directly. Compared with the traditional method of difference space (DS), RDS is reversible and it overcomes the ill-transformation problem. This method is applied to face recognition in Yale Face database B, and the recognition result demonstrates its robust performance under different illumination conditions
Currently in the community of medical imaging computation and analysis, there exist several excellent toolkits for various tasks. For example, VTK (http://***) is the most famous visualization tool and ITK (http://***...
详细信息
Currently in the community of medical imaging computation and analysis, there exist several excellent toolkits for various tasks. For example, VTK (http://***) is the most famous visualization tool and ITK (http://***) plays an important role in the segmentation and registration of medical images. These toolkits can assist the doctors to get the more accurate diagnosis, and help the researchers to get the more powerful capability of medical image processing. However, they are separate and style-various. Even for VTK and ITK, which are designed by the same company, users have to familiarize the two completely different coding styles. In this situation, using all of these toolkits is difficult. In this study, we will introduce 3D medical imaging computation and analysis platform (3DMed), which provides powerful functions such as image preprocessing, virtual cutting, surface rendering, volume rendering, manipulation and virtual endoscopy. The functional module description, system design and the visualization algorithm is demonstrated. The system can be widely applied to processing and analysis of CT and MR images.
In MR imaging, image noise, bias field, and partial volume effect are adverse phenomena that increases inter-tissue overlapping and hampers quantitative analysis. This study provides a powerful fully automated classif...
详细信息
ISBN:
(纸本)0769525210
In MR imaging, image noise, bias field, and partial volume effect are adverse phenomena that increases inter-tissue overlapping and hampers quantitative analysis. This study provides a powerful fully automated classification method, which combines the bias field correction and PV segmentation together. The method has been validated on simulated and real MR images for which gold standard segmentation available. The experimental results show that the proposed method is more accurate and robust than currently available models
Much attention has been paid upon network data flow control in recent years. The main problem in this field is how to design good algorithm or control law for flow rate of network data flow sources and for updated pri...
详细信息
暂无评论