检索结果-内蒙古大学图书馆

Spectral and spatial algorithm architecture for classification of hyperspectral and LIDAR

Proceedings of SPIE - The International Society for Optical Engineering 2012年 8407卷

作者： Rand, Robert S. Khuon, Timothy S. National Geospatial-Intelligence Agency Office of Innovision Basic and Applied Research 7500 GEOINT Drive Springfield VA 22153 United States

ISBN: (纸本)9780819490858

Architecture for neural net multi-sensor data fusion is introduced and analyzed. This architecture consists of a set of independent sensor neural nets, one for each sensor, coupled to a fusion net. The neural net of each sensor is trained (from a representative data set of the particular sensor) to map to a hypothesis space output. The decision outputs from the sensor nets are used to train the fusion net to an overall decision. To begin the processing, the 3D point cloud LIDAR data is classified based on a multi-dimensional mean-shift segmentation and classification into clustered objects. Similarly, the multi-band HSI data is spectrally classified by the Stochastic Expectation-Maximization (SEM) into a classification map containing pixel classes. For sensor fusion, spatial detections and spectral detections complement each other. They are fused into final detections by a cascaded neural network, which consists of two levels of neural nets. The first layer is the sensor level consisting of two neural nets: spatial neural net and spectral neural net. The second level consists of a single neural net, that is the fusion neural net. The success of the system in utilizing sensor synergism for an enhanced classification is clearly demonstrated by applying this architecture for classifying on November 2010 airborne data collection of LIDAR and HSI over the Gulfport, MS, area. © 2012 SPIE.

关键词： sensor data fusion

来源：评论

学校读者我要写书评

暂无评论

Simulation-based development of embedded sensor fusion applications

Simulation-based development of embedded sensor fusion appli...

引用

International Conference on Computational Cybernetics (ICCC)

作者： W. Elmenreich M. Schlager University of Technology Vienna Vienna Austria

This paper proposes a methodology for the design, testing, and fine-tuning of sensor fusion applications for distributed embedded sensor networks. The presented approach involves the following steps: (i) gathering data about the process environment, (ii) modeling of the process and the sensors to be used, (iii) selection of sensor fusion functions to process the sensor measurements, (iv) simulated closed-loop hardware-in-the-loop testing, (v) validation by open-loop testing in the real process, and, finally, (vi) application of the developed embedded sensor fusion application in the system. Particular steps may be repeated in a loop, if the result of a given step is not satisfactory. The main contribution of the approach is a separation of the modeling of process environment, sensors, and sensor processing. This allows for a strategic refinement of the model and supports the reuse and change of parts, e. g., in case of a sensor upgrade. The application of the approach is shown by a MATLAB/Simulink model that incorporates block diagrams describing process environment, sensor behavior, and sensor fusion algorithms

关键词： sensor fusion sensor systems Intelligent sensors Design methodology sensor systems and applications System testing Intelligent actuators Data processing sensor phenomena and characterization Partitioning algorithms

来源：评论

学校读者我要写书评

暂无评论

Resource-Efficient sensor fusion via System-Wide Dynamic Gated Neural Networks 21

Resource-Efficient Sensor Fusion via System-Wide Dynamic Gat...

引用

21st Annual IEEE International Conference on Sensing, Communication, and Networking, SECON 2024

作者： Singhal, C. Wu, Y. Malandrino, F. de Guevara Contreras, S. Ladron Levorato, M. Chiasserini, C.F. INRIA France UC Irvine United States CNR-IEIIT Italy CNIT Italy Politecnico di Torino Italy Chalmers University of Technology Sweden

ISBN: (纸本)9798331519186

Mobile systems will have to support multiple AI-based applications, each leveraging heterogeneous data sources through DNN architectures collaboratively executed within the network. To minimize the cost of the AI inference task subject to requirements on latency, quality, and - crucially - reliability of the inference process, it is vital to optimize (i) the set of sensors/data sources and (ii) the DNN architecture, (iii) the network nodes executing sections of the DNN, and (iv) the resources to use. To this end, we leverage dynamic gated neural networks with branches, and propose a novel algorithmic strategy called Quantile-constrained Inference (QIC), based upon quantile-Constrained policy optimization. QIC makes joint, high-quality, swift decisions on all the above aspects of the system, with the aim to minimize inference energy cost. We remark that this is the first contribution connecting gated dynamic DNNs with infrastructure-level decision making. We evaluate QIC using a dynamic gated DNN with stems and branches for optimal sensor fusion and inference, trained on the RADIATE dataset offering Radar, LiDAR, and Camera data, and real-world wireless measurements. Our results confirm that QIC matches the optimum and outperforms its alternatives by over 80%. © 2024 IEEE.

关键词： Constrained optimization

来源：评论

学校读者我要写书评

暂无评论

A Five Layer Architecture for Real Time Object Detection in Indoor Environments

引用

IFAC Proceedings Volumes 2000年第28期33卷 315-320页

作者： Hartmut Surmann German National Research Center for Information Technology Institute for Autonomous intelligent Systems Schloß Birlinghoven D-53754 Sankt Augustin Germany

The prognosis for autonomous mobile service robots for transportation tasks in indoor environments, e.g. multistory buildings differs a lot from todays reality. The robots have to act in dynamic environments with a huge number of components. The robots total repertoire of affordances and intelligence is highly according to the complexity of the building and its respective task. Difficult tasks can only be achieved of the base on immediate sensing of the environment and real-time evaluation of the sensor processing modules. This paper describes two components; a five layer sensor architecture which integrates a spatial database for multistory buildings and different modules of medium and limited intelligence but with real-time capabilities. A basic feature extraction algorithms of layer ii is presented as an example of the modules.

关键词： Programming and sensor architectures collision avoidance and sensor-based control robot sensing and data fusion

来源：评论

学校读者我要写书评

暂无评论

A new multi-sensor fusion algorithm based on the Information Filter framework

A new multi-sensor fusion algorithm based on the Information...

引用

Seminar on Detection Systems architectures and Technologies (DAT)

作者： Ammar Cherchar Messaoud Thameri Adel Belouchrani Ecole Supérieure Ali CHABATI Réghaia Algeria Ecole Nationale Polytechnique Algiers Algeria

ISBN: (纸本)9781509045099

his paper presents a new efficient track-to-track fusion (T2TF) algorithm based on the Information Filter (IF) framework which takes into account phenomena encountered in practical applications. In fact, it combines a modified version of the IF framework and a self-tuning fusion procedure based on likelihood functions to address issues such as the correlation of the estimates, the transmission shortcomings and the measurement origin uncertainty. The proposed method is evaluated through simulations and the obtained results show that the proposed algorithm performs as well the optimal centralized fusion schema in an idealized environment while it exhibits better robustness capabilities than existing decentralized algorithms when observation origin uncertainty is considered. Moreover, its reduced complexity cost is suitable for real time applications.

关键词： Microwave measurement Antenna measurements Microstrip antenna arrays Antenna radiation patterns Microwave antenna arrays

来源：评论

学校读者我要写书评

暂无评论

Smart environmental architecture for node localization in a wireless sensor network

Smart environmental architecture for node localization in a ...

引用

8th International Conference on Intelligent Environments, IE 2012

作者： Jauregui-Ortiz, Salvador Siller, Mario Ramos, Felix Scalabrin, Edson Electrical Engineering and Computer Science Department Center of Research and Advanced Studies Guadalajara Jalisco Mexico Graduate Program in Computer Science Pontifical Catholic University Parana Brazil Curitiba Parana Brazil

ISBN: (纸本)9780769547411

As far as we know, there is not a Node Localization Algorithm (NLA) that presents the same accuracy for all possible scenarios. We believe that a NLA should be able to interpret the dynamic information of the environment. In this sense, simple NLAs are rather focused and might perform well for specific scenarios and applications. Therefore, information fusion and context awareness seems to be an appropriate approach to address this issue. We propose the Smart Environmental Architecture for Node Localization (SEA-NL), which is composed by two main elements: (i) the Smart Beacon Nodes (SBNs) and (ii) the Logical Position of Nodes (LPN). In (i) the obstruction level indicator is estimated and can improve the estimation of distances among nodes. In (ii) environment information and a one to one relation between a node and an object are used and can also improve location estimation. Via simulation, our architecture was tested indoors and outdoors considering three localization algorithms: the Weighed Centroid Localization (WCL), the Centroid Localization, and the Triangular Centroid Localization. Finally, we present an accuracy comparison among NLAs used in isolated way, and by using the SBNs, the LPN, and the SEA-NL, where our architecture improves WCL up to ~30.88%. © 2012 IEEE.

关键词： Information fusion

来源：评论

学校读者我要写书评

暂无评论

RGB and LiDAR fusion based 3d semantic segmentation for autonomous driving

arXiv

引用

arXiv 2019年

作者： Madawi, Khaled El Rashed, Hazem Sallab, Ahmad El Nasr, Omar Kamel, Hanan Yogamani, Senthil Valeo R & D Cairo Egypt Cairo University Valeo Visions Systems Ireland

LiDAR has become a standard sensor for autonomous driving applications as they provide highly precise 3D point clouds. LiDAR is also robust for low-light scenarios at night-time or due to shadows where the performance of cameras is degraded. LiDAR perception is gradually becoming mature for algorithms including object detection and SLAM. However, semantic segmentation algorithm remains to be relatively less explored. Motivated by the fact that semantic segmentation is a mature algorithm on image data, we explore sensor fusion based 3D segmentation. Our main contribution is to convert the RGB image to a polar-grid mapping representation used for LiDAR and design early and mid-level fusion architectures. Additionally, we design a hybrid fusion architecture that combines both fusion algorithms. We evaluate our algorithm on KITTI dataset which provides segmentation annotation for cars, pedestrians and cyclists. We evaluate two state-of-the-art architectures namely SqueezeSeg and PointSeg and improve the mIoU score by 10% in both cases relative to the LiDAR only baseline. Copyright © 2019, The Authors. All rights reserved.

关键词： Autonomous vehicles

来源：评论

学校读者我要写书评

暂无评论

Data fusion in Wireless sensor Network: An Overview

Data Fusion in Wireless Sensor Network: An Overview

引用

CHILEAN Conference on Electrical, Electronics Engineering, Information and Communication Technologies (CHILECON)

作者： Daniel Fernando Quintero Bernal Héctor Kaschel John Kern Departamento de Ingeniería Eléctrica Universidad de Santiago de Chile Santiago Chile

Due to the massive increase in data sets in contemporary applications and the difficulty of their processing in various approaches, data fusion remains a predominant way to obtain outstanding results, in terms of reliability, efficiency, and precision. The following document presents an overview of data fusion in wireless sensor networks; in which, various concepts to applications are considered to generate a base document for future research. This article gives a comprehensive and current view of models or architectures, techniques, methodologies or algorithms, and a current review of applications in different areas.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Data fusion in multi sensor platforms for wide-area perception

Data fusion in multi sensor platforms for wide-area percepti...

引用

IEEE Symposium on Intelligent Vehicle

作者： A. Polychronopoulos N. Floudas A. Amditis D. Bank B. van den Broek Institute of Communications and Computer Systems Athens Greece Daimler Chrysler Aerospace Ulm Germany TNO Defensie en Veiligheid Den Haag Netherlands

There is a strong belief that the improvement of preventive safety applications and the extension of their operative range are achieved by the deployment of multiple sensors with wide fields of view (FOV). The paper contributes to the solution of the problem and introduces distributed sensor data fusion architectures and algorithms for an efficient deployment of multiple sensors that give redundant or complementary information for the moving objects. The proposed fusion architecture is based on a modular approach allowing exchangeability and benchmarking using the output of individual trackers, whereas the fusion algorithm gives a solution to the track management problem and the coverage of wide perception areas. The test case is LATERAL SAFE sensor configuration, which monitors the rear and lateral areas of the vehicle. Results show that with the given approach the system is able to maintain the ID of all objects in transition (an object enters a sensor's FOV) and blind areas (no sensor coverage)

关键词： sensor fusion sensor systems Laser radar Infrared sensors Vehicle safety Accidents sensor systems and applications Radar tracking Intelligent sensors Benchmark testing

来源：评论

学校读者我要写书评

暂无评论

Parallel Computing architectures for Robotic applications: A Comprehensive Review

arXiv

引用

arXiv 2024年

作者： Islam, Md Rafid Electrical and Computer Engineering North South University Dhaka Bangladesh

With the growing complexity and capability of contemporary robotic systems, the necessity of sophisticated computing solutions to efficiently handle tasks such as real-time processing, sensor integration, decision-making, and control algorithms is also increasing. Conventional serial computing frequently fails to meet these requirements, underscoring the necessity for high-performance computing alternatives. Parallel computing, the utilization of several processing elements simultaneously to solve computational problems, offers a possible answer. Various parallel computing designs, such as multi-core CPUs, GPUs, FPGAs, and distributed systems, provide substantial enhancements in processing capacity and efficiency. By utilizing these architectures, robotic systems can attain improved performance in functionalities such as real-time image processing, sensor fusion, and path planning. The transformative potential of parallel computing architectures in advancing robotic technology has been underscored, real-life case studies of these architectures in the robotics field have been discussed, and comparisons are presented. Challenges pertaining to these architectures have been explored, and possible solutions have been mentioned for further research and enhancement of the robotic applications. © 2024, CC BY.

关键词： Decision making

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：