基于双镜头视野协同成像的无线视频传感器网络构建

Construction of Wireless Video Sensor Network Based on Dual Lens Field of View Collaborative Imaging

摘要: 针对地处边远地区江河湖库超大水域观测、大坝工程瞭望、恶劣天气条件下水面漂浮物检测、异常事件辨识及水体水质视觉监测等水利领域智能视频监控特有的应用需求，通过改进无线视频传感器网络(WVSN)、仿“复眼”成像以及视频边缘计算等现有的技术方法，本文提出一种基于双镜头视野协同成像的无线视频传感器网络(简称：WVSN-CI)设计新模式，为广域大范围的全域性水利智能视频监控系统建设提供前端网络设计解决方案。本文还对双镜头摄像机节点组成的WVSN-CI关键技术做出介绍，包括介绍和分析人工复眼与常规CMOS图像传感器双镜头合成，双镜(头)三核(芯)摄像机节点设备的光–机–电–算一体化设计，WVSN-CI视频数据传输协议，基于视频边缘计算的汇聚节点，以及WVSN-CI与水利信息化基础网络设施和视频主控中心的链接等内容。

Abstract: Aiming at the special application needs of intelligent video surveillance in the water conservancy field, such as observation of super large waters of rivers, lakes and reservoirs in remote areas, observation of dam projects, detection of floating objects on the water surface under bad weather conditions, identification of abnormal events and visual monitoring of water quality, by improving existing technical methods such as wireless video sensor network (WVSN), imitation compound eye imaging and video edge computing, this paper proposes a new design pattern for wireless video sensor networks based on dual lens field of view collaborative imaging (referred to as WVSN-CI), that the pattern can provide an front-end network design solutions for the construction of wide area and wide range intelligent video surveillance system in water conservancy. The paper also introduces and analyzes the key technologies of the WVSN-CI composed of dual lens camera nodes, including dual lens synthesis of artificial compound eye and conventional CMOS image sensors, the integrated design of optical, mechanical, electrical, and computing of “dual mirror -three core” camera node equipment, video data transmission protocol for WVSN-CI, sink node based on video edge computing, and links between WVSN-CI and water conservancy information infrastructure, video control center, etc.