【作者】 唐文；

【导师】 蒋平； 张伟；

【作者基本信息】 同济大学 ， 控制理论与控制工程， 2007， 硕士

【摘要】 本文在研究现有无线传感器网络节点硬件平台特点的基础上，设计并实现了一种利用高性能ARM7处理器，支持IEEE 802．15．4协议的CC2420芯片和三维振动，温度等多种传感器搭建的新型无线传感器节点硬件平台OpenWSN。OpenWSN能够承担图像采集与传输、汇聚节点等对性能要求更高的应用，可与目前基于MCU方式的节点形成良好的互补。设计遵循了节点的标准化、微型化，低功耗和扩展性等原则。对该节点的测试表明该方案较好地实现了上述目标，因此它可以成为研究人员选用无线传感器网络平台的理想高端选择。目前，该传感器网络平台应用在智能环境中的机器人导航研究中，每个传感器节点携带一个低分辨率摄像头以感知周边环境及其变化，覆盖广大面积的多套节点将借助无线通信联合工作，提供针对环境的较完备信息并可支持在网处理。节点设计中引入了FPGA芯片以硬件加速方式支持快速图像处理。FPGA+摄像头+OpenWSN称为WiME。另外，论文还研究并提出了一种利用多个低分辨率图像以提高场景图像分辨率的方法。该方法通过模板获得图像间的单应矩阵，然后通过图像间的单应矩阵将图像映射到另一图像后，将不规则图像样本点以一定的方法进行不规则数据插值。该方法将为智能环境中的机器人及其导航提供可信的环境信息。更多还原

【Abstract】 This paper proposes and implements a new open research platform of Wireless sensor networks (WSN) named OpenWSN. The scheme utilizes microcontroller of ARM family, wireless RF transceiver accorded with 802.15.4 specification, three dimension vibration sensor, temperature sensor and etc. Larger sized than Telos, OpenWSN has higher processing capability than current pervasive research hardware like Gains, Mica and Telos. OpenWSN can bear higher performance application, such as image data collection and pan coordinator, thus preferably complement current MCU scheme. OpenWSN follows the principle of standardization, miniature low energy consumption and expansibility. Test of the node shows that the implementation achieves the goal preferably and is an ideal choice of wireless sensor network platform for researchers.At present, the platform of wireless sensor network is used in robot navigation directed by intelligent environment. Every node carries a low resolution camera to perceive environment and its modification. Covering vast area, a great many nodes cooperate in wireless communication mode and provide enough information about environment and also network process. The node introduces FPGA to support fast image process by hardware acceleration mode. Camera, FPGA and OpenWSN is called WiME.In addition, the paper proposes one approach to enhance image resolution by a series low resolution images got by low resolution cameras. The approach is based on homography between two planes calculated by image template and the method maps one image into another image and then does interpolation under scatter data to achieve resolution enhancement.更多还原