A novel image deblurring method based on high-order non-local range Markov Random Field (NLR-MRF) prior is proposed in the paper. NLR-MRF is an effective statistical framework to model prior knowledge of natural image...
详细信息
A novel image deblurring method based on high-order non-local range Markov Random Field (NLR-MRF) prior is proposed in the paper. NLR-MRF is an effective statistical framework to model prior knowledge of natural images which leads to excellent performance in some low-level vision problems. In our work, the framework is extended to image deblurring. To overcome some limitations of maximum a-posteriori (MAP) estimation, we adopt Bayesian minimum mean squared error (MMSE) estimation to perform deblurring. The high-order NLR-MRF prior can be easily integrated into this framework. Then, an efficient Gibbs sampling algorithm is employed to compute MMSE estimation. The proposed method frees the user from determining regularization parameter beforehand, which relies on unknown noise level. Our deblurring method shows superior or comparable results to the state-of-art deblurring methods.
This paper illustrates a compositional deformable model for detecting vehicle and recognizing vehicle-contours. To overcome the difficulties that vehicles in an image have various sizes, shapes, colors and poses, this...
详细信息
This paper illustrates a compositional deformable model for detecting vehicle and recognizing vehicle-contours. To overcome the difficulties that vehicles in an image have various sizes, shapes, colors and poses, this model has two main characteristics: first, the model is made up of constituent parts which shared by vehicles. The locality of parts give the model the ability to recognize vehicles with different types (e.g., although vehicles have various sizes and shapes, they are usually composed by roof, windscreen, windows, etc.). Second, the spatial relationships of these parts are represented by Markov Random Field (MRF). The model is deformable to adapt to vehicles of different shapes and poses because of the appropriately changing of combinations of these parts in the MRF. Experimental results with real world images show that this method is effective in vehicle detection and vehicle-contours recognition.
Road traffic load balancing can avoid network congestion and improve traffic efficiency. This paper proposes a method of dynamic route guidance based on Maximum Flow Theory to balance traffic load of road network. A m...
详细信息
Road traffic load balancing can avoid network congestion and improve traffic efficiency. This paper proposes a method of dynamic route guidance based on Maximum Flow Theory to balance traffic load of road network. A modified Ford-Fulkerson algorithm is used for searching the optimal route. In addition, the algorithm is implemented by using MapReduce primitives, which introduces Cloud Computing Platform for large-scale traffic network guidance. Computational experimental environment is built by integrating Artificial Transportation systems (ATS) and Hadoop. Results in ATS and performances on Hadoop show that the method proposed can improve the traffic situation effectively.
Radio fingerprint matching localization method promises high localization accuracy but requires extensive infrastructural effort and search operations. This paper proposes a new search strategy for radio fingerprint m...
详细信息
Radio fingerprint matching localization method promises high localization accuracy but requires extensive infrastructural effort and search operations. This paper proposes a new search strategy for radio fingerprint matching method, which sharply reduces search operations and shows little effect on localization accuracy. Indoor experiment is conducted to evaluate our method.
It is important for a chemical plant to find a suitable performance appraisal method. In this paper, based on the ACP (artificial system, computational experiment, and parallel execution) theory and the PageRank algor...
详细信息
It is important for a chemical plant to find a suitable performance appraisal method. In this paper, based on the ACP (artificial system, computational experiment, and parallel execution) theory and the PageRank algorithm, a new performance appraisal method is proposed. The proposed method comprehensively involves both peoples and routine management rules from the holistic viewpoint. By comparison with the traditional performance appraisal method, the proposed method is more reasonable, more flexible and robuster.
Chemical industry is complex and continuous process industry, and the control and management of the long-term safe operation involves a great deal of information and data on the staffs, management, equipment and techn...
详细信息
Chemical industry is complex and continuous process industry, and the control and management of the long-term safe operation involves a great deal of information and data on the staffs, management, equipment and technology. Consequently, making real-time assessment for the control and management by computational experiment is important for the long-term of safety operation of chemical industry. As the requirement for high-performance computing in large-scale complexsystems, traditional CPU computing can not meet the needs of large-scale computing. Because of the high speed and rapid development features, GPU was widely applied to scientific computation, which is suitable for large-scale intensive computation. This paper carried on specific computation experiments using GPU parallel computing, and designed and analyzed matrix inversion, fuzzy logic computation and Monte Carlo computation respectively for staff management and equipment simulation. Finally, compared to CPU serial computing, the maximum speedup ratio, 34, 85 and 39 times of speedup ratio are achieved.
Learning control has been an active topic of research for several decades, and is of theoretical, as well as practical, significance. Current theories and developments in learning control are discussed. Following ...
详细信息
Learning control has been an active topic of research for several decades, and is of theoretical, as well as practical, significance. Current theories and developments in learning control are discussed. Following a brief introduction of the state as well as new progress on learning control, we give a detail review on the models and algorithms of the control policies developed recently which proved to be advantageous over previous approaches through experimental results. The related results and properties are presented. Then, several potentially developmental topics that are valuable to be further investigated are suggested. Finally, the conclusion remark is proposed.
Two humanoid robots are used to play table tennis with each other. For each humanoid robot, three cameras and a computer are equipped to form a stereovision system and a monocular vision system. The stereovision syste...
详细信息
Two humanoid robots are used to play table tennis with each other. For each humanoid robot, three cameras and a computer are equipped to form a stereovision system and a monocular vision system. The stereovision system consisting of two smart cameras and a computer measures the 3-dimensional position of the table tennis ball. It adopts parallel processing mode in order to realize hundred frames level measurement per second. A high-speed digital camera and the computer compose the monocular vision system, which measures the pose of the robot relative to the table via a color mark attached on the robot. The two smart cameras in each stereovision system are synchronized via I/O signals. The vision systems for the two robots are synchronized by time verification. Experimental results verify the effectiveness of the designed vision system and the proposed methods.
The present work focuses on the node deployment algorithm of Wireless Sensor Networks. The Central Voronoi Tessellation algorithm is employed to optimize the node position. The energy consumption of the whole sensor n...
详细信息
The present work focuses on the node deployment algorithm of Wireless Sensor Networks. The Central Voronoi Tessellation algorithm is employed to optimize the node position. The energy consumption of the whole sensor network will be minimized by using this algorithm. Simulation of the proposed algorithm shows the effectiveness of minimizing the energy consumption.
An analytic flying model that can well represent the physical behavior is derived, where the ball's self-rotational velocity changes along with the flying velocity. Based on the least square method, a rebound mode...
详细信息
An analytic flying model that can well represent the physical behavior is derived, where the ball's self-rotational velocity changes along with the flying velocity. Based on the least square method, a rebound model that represents the relation between the velocities before and after rebound is established. The initial trajectory is fitted to three second order polynomials of the flying time with the measured positions of the ball. The initial velocities of the ball in the analytic flying model, including the flying velocity and the self-rotational velocity, are computed from the polynomials. The ball's landing position and velocity is predicted with the model. The velocities after rebound are determined with the rebound model. By taking the velocities after rebound as new initial ones, the flying trajectory after rebound is described with the model again. In other words, the ball's trajectory is predicted. Experimental results verify the effectiveness of the proposed method.
暂无评论