航拍图像的无线传感网节点识别与定位算法

为提高节点定位精度并减少能量耗费，提出一种基于航拍图像处理的节点识别与定位算法.该算法主要思想是将基于深度学习模型的图像识别技术应用到节点识别与定位中.首先通过一个携带GPS的航拍器在节点部署区域自主采集节点图像，利用图像识别技术精确识别定位航拍器所拍摄节点，然后利用贝叶斯平均模型为已识别节点匹配节点编号.最后，将被定位节点升级为锚节点，结合经典定位算法确定未被拍摄的节点位置.通过大量的仿真证明了该算法的有效性.此外，利用25 Micaz节点和一个航拍器进行了实验，实验结果表明，该算法具有较好的鲁棒性、扩展性及定位精度，与传统定位算法相比定位精度提高约10%~30%.

Node Identification and Localization Algorithm Based on Aerial Image for Wireless Sensor Network

In order to improve localization accuracy and reduce energy consumption, a node identification and localization algorithm based on aerial image processing is proposed in this paper. The main idea of the algorithm is to apply the image recognition technology with deep learning model for node identification and localization. First, node images are collected autonomously through an aerial device carrying GPS in the deployment region. Second, image recognition technology is used to identify the nodes accurately. Finally, the identified nodes are updated to anchor nodes, and the classical location algorithms are adopted to determine the node locations that have not been photographed. Extensive simulations are conducted to verify the algorithm effectiveness. Moreover, the design is implemented with 25 Micaz nodes and an aerial photography. Test results show that the proposed algorithm has better robustness, expansibility and accuracy of positioning. Compared with the traditional localization algorithm, the positioning accuracy is improved about 10%~30%.