无人机河湖巡检中RCNN识别算法应用研究

Research on application of RCNN identification algorithmin UAVS river and lake patrol inspection

随着经济发展与人口增长,带来了河湖治理难度增加、水体污染频发等问题,针对这些情况设计出一种基于区域卷积神经网络算法(RCNN)的无人机河湖巡检图像智能处理平台。通过聚类和构建卷积神经网络提取可能含有河湖污染的区域,利用RCNN对目标区域进行分类和识别,以此作为问题源构建河湖管理的闭环,较好的完成巡检目标。

With economic development and population growth,problems such as increasing difficulty in river and lake control and frequent water pollution have been brought about.In response to these situations,an intelligent processing platform for UAV river and lake inspection images based on regional convolutional neural network algorithm(RCNN)was designed.Through clustering and constructing a convolutional neural network,areas that might contain river and lake pollution were extracted,and RCNN was used to classify and identify the target area,which was used as the source of the problem to construct a closed loop of river and lakE-management,and to better complete the inspection target.