GUAN, M Xinyuan (2023) Inspection de Construction avec les Drones Bon Marchés PRE - Research Project, ENSTA.

[img]PDF
Restricted to Registered users only

935Kb

Abstract

With the rapid development of robot technologies, the UAVs show a great potential in performing construction inspection tasks with low cost. Several papers have developed inspection AI algorithms for the UAVs with limited resources, but none of them built a general system for automatic UAV inspection tasks. In this project, we proposed a UAV inspection system composed of two parts: an efficient inspection algorithm to detect cracks and a route planning algorithm to let the UAV scan the wall in optimized time with optimized energy consumption. The system is tested in the simplified case where there is only one single wall to inspect, and the results showed that it achieved a satisfying performance in crack detection and can produce an optimized plan for wall scanning. The proposed system can provide a prototype for the future UAV inspection systems of real construction sites.

Item Type:Thesis (PRE - Research Project)
Uncontrolled Keywords:Inspection, UAV, neural network
Subjects:Information and Communication Sciences and Technologies
ID Code:9704
Deposited By:Astrid Auzou-Connes
Deposited On:01 sept. 2023 09:14
Dernière modification:01 sept. 2023 09:14

Repository Staff Only: item control page