Keywords
Deep Learning, Reinforcement Learning, Image Processing, Drone, Autonomous
Select the category the research project fits.
Innovative Technology/Entrepreneurship/Design
Is this submission part of ICaP/PW (Introductory Composition at Purdue/Professional Writing)?
No
Abstract
The objective of the group is to investigate the application of Reinforced Deep Learning to autonomously complete damage inspection on buildings after natural disasters. By using Reinforced Deep Learning with Convolutional Neural Networks performing image processing, the drone will be trained to navigate itself towards buildings and perform building inspection without scanning the whole structure. The algorithms for Reinforced Deep Learning can be used to drive drones autonomously and perform inspections currently done by humans. The stability of the structure of buildings are unknown when these inspections are performed. This application would greatly reduce the risk of individuals involved in inspecting the building structures and provide greater insight into the stability of these structures.
Recommended Citation
Biswas, Romita; Friedrich, Cassidy; Rao, Yug N.; and Chakraborty, Aditya, "Deep Reinforcement Learning for Autonomous Inspection System" (2019). Purdue Undergraduate Research Conference. 29.
https://docs.lib.purdue.edu/purc/2019/Posters/29
Deep Reinforcement Learning for Autonomous Inspection System
The objective of the group is to investigate the application of Reinforced Deep Learning to autonomously complete damage inspection on buildings after natural disasters. By using Reinforced Deep Learning with Convolutional Neural Networks performing image processing, the drone will be trained to navigate itself towards buildings and perform building inspection without scanning the whole structure. The algorithms for Reinforced Deep Learning can be used to drive drones autonomously and perform inspections currently done by humans. The stability of the structure of buildings are unknown when these inspections are performed. This application would greatly reduce the risk of individuals involved in inspecting the building structures and provide greater insight into the stability of these structures.