无线多媒体传感器网络论文：基于无线多媒体传感器网络的图像分布式处理技术的研究.doc

《无线多媒体传感器网络论文：基于无线多媒体传感器网络的图像分布式处理技术的研究.doc》由会员分享，可在线阅读，更多相关《无线多媒体传感器网络论文：基于无线多媒体传感器网络的图像分布式处理技术的研究.doc（7页珍藏版）》请在淘文阁 - 分享文档赚钱的网站上搜索。

1、 无线多媒体传感器网络论文：基于无线多媒体传感器网络的图像分布式处理技术的研究【中文摘要】无线多媒体传感器网络是在传统的无线传感器网络基础上引入音频、视频、图像、声音等大容量信息的新型传感器网络,在智能家居、战场监控、交通监控等领域具有广阔的应用前景。无线多媒体传感器网络将传统传感器网路的自组织、无人值守等优点和新兴多媒体技术感知媒体丰富的特点有机结合起来,既具有传统传感器网络(WSNs)的自组织、多跳路由、资源受限等共同特点,又在能耗分布、服务质量(QoS)要求、网络安全、传感器模型等方面具有显著的个性化特点。由于无线多媒体传感器网络中节点配备电池供电和低功耗的处理芯片,它的能量、处理能力、

2、存储能力、通信能力等都十分受限,使得能量的高效利用成为无线传感器网络的首要目标。在无线多媒体传感器网络中图像信息提供更精确丰富的监测信息,但是信息量巨大,因此如何有效的压缩和传输图像至关重要。在无线多媒体传感器网络(WMSNs)中,通常使用数据压缩编码、冗余信息过滤以及数据融合等处理技术减少数据的传输量,考虑到WMSNs单个节点能量有限和处理能力有限的特点,可以利用多个节点的协同处理能力,通过“在网计算”完成数据压缩编码等处理,将冗余度低、数据量小、信息量足的处理结果逐级上传至控制中心。针对WMSNs中图像处理的特殊要求,本文在研究WMSNs的分布式图像压缩算法的基础上,对已有的算法改进,提出

3、了一种适用于WMSNs的自适应分布式图像压缩算法；此外对当前的分布式视频编码技术进行了研究,探讨了其在图像压缩方面的应用。首先考虑到Walsh-Hadamard变换简单高效和能量集中的特点,可以将该变换算法应用于图像压缩算法中。由于Walsh变换采用二值正交基,非常适合计算机实现,而且有相应的快速计算机算法实现。针对WMSNs节点处理能力有限的特点,这种变换较DCT变换、小波变换需要更少的存储空间,提高了运算速度。仿真实验结果表明采用Walsh-Hadamard变换的算法与基于DCT变换算法相比更加简单快速。其次考虑到WMSNs无线信道的特点,提出一种基于信道估计的自适应图像压缩机制。无线信道

4、具有信号衰减,延迟大,高误码率突发数据传输等特点,而多媒体信息处理需要Qos保障,如延迟容忍度,数据传输所需带宽。可以基于信道状态的准确估计,自适应调整量化表,在保证图像质量满足应用的前提下,获得更高的压缩比率。最后对分布式信源编码理论和实用编码技术进行了研究,针对WMSNs的应用进行了讨论,探讨了其在图像压缩方面的应用。【英文摘要】WMSNs is on the basis of WSNs with introducing audio, video, image,sound, large-capacity information.Its widely used in environmenta

5、l monitoring, battlefield surveillance, traffic monitoring,ect.Besides self-organizing,unattended it also combines perception of media-rich multimedia technology. On the one hand,it is traditional self-organizing,multi-hop routing, resource constraints and other common characteristics.On the other h

6、and,the energy distribution,QoS requirments,sensor model and also has a significant personal characteristics.As energy of nodes, processing power,storage capacity of WMSNs isvery limited,efficient use of energy is the primary objective of WMSNs.The compression and transmission of image is critical.C

7、ompression,filtering,and data fusion of information processing technology are widely used in WSNs. Taking account of the limited energy of single node,processing power,we can use the processing power of multiple nodes, with “in network computing” we can complete the compression of multimedia informa

8、tion and upload the small amount of data with sufficient information to the next target step by step.Because of the special requirements,based on the improvement of existing algorithms,a distributed image compression algorithm is proposed.First because the Walsh-Hadamard transform is simple,efficien

9、t and enengy focused,so it can be applied to the image compression algorithm.lt is very suitable for computer implementation.Compared to DCT transfom and wavelet transform,it requires less storage space and improves the speed of operation.The simulation results show that the algorithm with Walsh-Had

10、amard transform is more simple and rapid than DCT-based algorithm.Secondly taking account of the characteristics of wireless channel in WSNs,we propose an adaptive image compression mechanism with channel estimation.Bacause wireless channel is signal-attenuation,large delay,high bit err rate and bur

11、st data transmission,multimedia information processing needs proper Qos,such as delay tolerance,data bandwidth,ect.Based on accurate channel state estimation,we can use adaptive quantization table, ensure that the quality of image meet the needs and get higher compression ratio.Simulation results sh

12、ow that the algorithm in intensive WMSNs can prolong the network life cycle.Finally according to the previous improvement we proposed the adaptive distributed image compression algorithm with Walsh-Hadamard transform.In the transform stage, the Walsh-hadamard tranfrom is used to reduce the node proc

13、essing overhead. In the quantitative phase,we use the channel estimation mechanism to quantilize adaptively and make higher compression ratio.The simulation results show that the proposed algorithm with limited resources and deployment of intensive WMSNs nodes,it prolongs the network lifetime signif

14、icantly.【关键词】无线多媒体传感器网络 Walsh-Hadamard变换 DCT变换 自适应图像压缩 分布式【英文关键词】wireless multimedia sensor network Walsh-Hadamard transform DCT transform adaptive image compression distributed compression【目录】基于无线多媒体传感器网络的图像分布式处理技术的研究摘要10-12ABSTRACT12-13第一章 绪论14-191.1 研究背景14-151.2 研究的目的和意义15-161.3 国内外研究现状16-181.4 本

15、文组织结构18-19第二章 无线多媒体传感器网络综述19-322.1 无线多媒体传感器网络概述19-232.1.1 无线多媒体传感器网络的概念19-202.1.2 无线多媒体传感器网络的特点20-222.1.3 无线多媒体传感器网络的应用22-232.2 无线多媒体传感器网络的关键技术及其研究进展23-312.2.1 节点系统23-242.2.2 MAC协议24-262.2.3 路由协议26-282.2.4 多媒体信息处理28-292.2.5 其它关键技术29-312.3 本章小结31-32第三章 基于Walsh-Hadamard变换的分布式图像压缩算法32-453.1 概述32-333.1.

16、1 无线多媒体传感器网络图像处理面临的问题323.1.2 有的解决方案及其局限性32-333.2 Walsh-Hadamard变换33-363.2.1 Walsh-Hadamard变换的概念33-343.2.2 Walsh-Hadamard变换的特点34-353.2.3 Walsh-Hadamard变换应用于图像压缩35-363.3 基于Walsh-Hadamard变换的分布式图像压缩算法36-413.3.1 系统模型36-373.3.2 基于Walsh-Hadamard变换的分布式图像压缩算法37-383.3.3 算法性能仿真38-413.4 基于Walsh变换的分布式图像压缩算法在tiny

17、os的设计与实现41-433.4.1 Tinyos简介41-423.4.2 基于walsh变换的图像压缩算法的概要设计42-433.5 本章小结43-45第四章 一种基于信道估计的自适应量化机制45-514.1 概述454.2 无线信道模型45-464.2.1 无线信道的特点454.2.2 无线信道模型45-464.3 一种基于信道估计的自适应量化机制46-494.3.1 自适应量化机制46-474.3.2 仿真验证47-494.4 自适应量化机制在tinyos操作系统下的概要设计49-504.4.1 协议栈模型设计494.4.2 信道评价器组件接口设计49-504.5 本章小结50-51第五

18、章 分布式信源编码理论与实用编码技术研究51-605.1 分布式信源编码理论51-555.1.1 引言51-525.1.2 Slepian-Wolf无损编码理论52-535.1.3 Wyner-Ziv有损编码理论53-555.2 实用编码技术55-595.2.1 引言555.2.2 DISCUS编码和基于Turbo码的编码技术55-575.2.3 分布式信源编码在WMSNs中图像压缩的应用57-595.3 本章小结59-60第六章 总结与展望60-626.1 结论606.2 展望60-62参考文献62-66致谢66-67攻读学位期间发表的学术论文目录67-68学位论文评阅及答辩情况表68act

19、ively carry out the law on civil air defense education, drawn out of the air defense in Pingliang city Building under easy fare, daily special inspection and regulation, overfulfilled the province upper and lower knots of up to 500,000 yuan fee collection tasks. 5, further standardize internal management, improve staff quality. Adhere to the