Many image completion methods are based on a low-rank approximation of the underlying image using matrix or tensor decomposition models. In this study, we assume that the image to be completed is represented by a mult...
详细信息
Many image completion methods are based on a low-rank approximation of the underlying image using matrix or tensor decomposition models. In this study, we assume that the image to be completed is represented by a multi-way array and can be approximated by a conical hull of subtensors in the observation space. If an observed tensor is near-separable along at least one mode, the extreme rays, represented by the selected subtensors, can be found by analyzing the corresponding convexhull. Following this assumption, we propose a geometric algorithm to address a low-rank image completion problem. The extreme rays are extracted with a segmented convex-hull algorithm that is suitable for performing noise-resistant non-negative tensor factorization. The coefficients of a conical combination of such rays are estimated using Douglas-Rachford splitting combined with the rank-two update least-squares algorithm. The proposed algorithm was applied to incomplete RGB images and a hyperspectral 3D array with a large number of randomly missing entries. Experiments confirm its good performance with respect to other well-known image completion methods.
Presently Wireless Sensor Network used to deploy at various critical locations where heavy traffic load situations along with priority-based data routing are always occurred. Under these situations, if any Wireless Se...
详细信息
Presently Wireless Sensor Network used to deploy at various critical locations where heavy traffic load situations along with priority-based data routing are always occurred. Under these situations, if any Wireless Sensor Network (WSN) node wants to send the emergency data to sink/server, then that node may face harsh traffic load conditions in various routes, therefore many times end-to-end delay occurs while sending emergency data. To find the best optimized possible solution for this problem, we proposed an intelligent self-learning and educating algorithm, which uses reinforcement learning technique, shortest path algorithm along with convex-hull algorithm for creation of safe path (green corridor) to send data packets from any node to sink/server in emergency situation on priority basis. This approach is beneficial for engineers (industry/students) to better deal with real-time emergency situations and education purpose. Mainly our proposed learning algorithm selects certain WSN nodes based upon their link quality and higher residual energy to create safe path (green corridor) for emergency situation. The performance of this learning algorithm has been tested in MATLAB and Contiki Cooja simulator under heavy traffic conditions along with various performance parameters like Packet Delivery Ratio (PDR) ratio, end-to-end delay, energy consumption and better throughput. Final result shows that our proposed learning algorithm Wireless Sensor Network - Reinforcement Learning (WSN-RL) performs better than existing Reinforcement Learning based lifetime optimization (RLLO) and Reinforcement Learning based Communication Range Control (RL-CRC) algorithms to achieve the optimum solution for our problem.
The measurement of total catch on-board a fishing vessel is generally a very complex process, especially to scientific observers. In this context, hauls of large volume, intricate fishing operations and limited access...
详细信息
The measurement of total catch on-board a fishing vessel is generally a very complex process, especially to scientific observers. In this context, hauls of large volume, intricate fishing operations and limited access to the capture increase this problem. In this paper, we propose a methodology to address the estimation of catch volume in codends of a crustacean fishery through LiDAR (light and radar) technology. A sensor was used to acquire a three-dimensional representation of an object located at a fixed distance from the device, thereby simulating a fishing codend. Then, a convex-hull algorithm was applied to this representation to obtain an estimation of its volume. Additionally, to obtain further insights, an experimental laboratory setup was used to emulate the volume estimation of catches on a fishing vessel. The dataset acquired by the system was subsequently analyzed to study the percentage errors associated with the estimation process and to test whether the selected variables are significant. The results indicate that there is considerable uncertainty related to the volume estimation, but it can be addressed using a statistical model. This work constitutes the first attempt to provide a methodology to estimate the catch volume of a codend in a Chilean fishery by generating new measurement alternatives for fishery monitoring programs, enforcement and management institutions, as well as the fishing industry.
Perimeter protection aims at identifying intrusions across the temporary base established by army in critical regions. convex-hull algorithm is used to determine the boundary nodes among a set of nodes in the network....
详细信息
Perimeter protection aims at identifying intrusions across the temporary base established by army in critical regions. convex-hull algorithm is used to determine the boundary nodes among a set of nodes in the network. To study the effectiveness of such algorithm, we opted three variations, such as distributed approach, centralized, and mobile approach, suitable for wireless sensor networks for boundary detection. The convex-hull approaches are simulated with different node density, and the performance is measured in terms of energy consumption, boundary detection time, and accuracy. Results from the simulations highlight that the convex-hull approach is effective under densely deployed nodes in an environment. The different approaches of convex-hull algorithm are found to be suitable under different sensor network application scenarios.
Given a set R of red points and a set B of blue points, the nearest-neighbour decision rule classifies a new point q as red (respectively, blue) if the closest point to q in R boolean OR B comes from R (respectively, ...
详细信息
Given a set R of red points and a set B of blue points, the nearest-neighbour decision rule classifies a new point q as red (respectively, blue) if the closest point to q in R boolean OR B comes from R (respectively, B). This rule implicitly partitions space into a red set and a blue set that are separated by a red-blue decision boundary. In this paper we develop output-sensitive algorithms for computing this decision boundary for point sets on the line and in R-2. Both algorithms run in time O(n log k), where k is the number of points that contribute to the decision boundary. This running time is the best possible when parameterizing with respect to n and k.
暂无评论