Open Access
Journal Article
Design of unmanned aerial vehicle system for intelligent inspection of high-rise buildings
by
Junjie Chen
, Naru Yue
, Jiayi Li
, Xu Wang
and
Xinzhuang Li
Abstract
In order to realize the purpose of better inspection on the outer surface of high-rise buildings, comprehensively detect the damage of the building surface, so as to find various problems, such as cracks, bulge and falling off, a UAV inspection image acquisition and processing system based on convolutional neural network is designed. The UAV uses open MV to realize the ef
[...] Read more
In order to realize the purpose of better inspection on the outer surface of high-rise buildings, comprehensively detect the damage of the building surface, so as to find various problems, such as cracks, bulge and falling off, a UAV inspection image acquisition and processing system based on convolutional neural network is designed. The UAV uses open MV to realize the efficient processing of visible light image information, and uses an embedded system for image fusion processing. Using convolutional neural network model and multi-object detection algorithm, extract the feature value of broken position and combine multiple sensors for 3 D position transformation. Data reads and 3 D coordinate map presentation were performed using the app software. For the safety problem of ceramic tiles, a marking device is installed on the UAV. Calculate the threat range and use the application to send location information to maintenance personnel and alert information to residents, and provide real-time location safety monitoring. The experimental results show that the designed uav can realize comprehensive, automatic, intelligent, efficient and preventive inspection of the building health status and problems of buildings through inspection, data collection and processing, intelligent analysis and early warning.