1. Damage assessment using image processing and convolutional neural networks.
- Author
-
Kaur, Loveneet, Bansal, Pranav, and Rani, Ekta
- Subjects
CONVOLUTIONAL neural networks ,IMAGE processing ,NATURAL disasters ,DAMAGE claims - Abstract
Damage assessment is one of the main challenge for the various insurance organizations and other agencies with the task to access damage for financial claims or to make prediction of renovation cost. This study focuses to observe images acquired from internet or satellites for the utilization of distinguishing and dealing with a natural calamity. Images taken from satellite or air drone can depicts the state of ground and the data further could be analyzed for prediction of damage. This paper proposes a methodology to detect the flood affected area with hybrid approach using wavelet and convolutional neural network. This approach will further helps to assess the damage the by classifying the type of damage on the basis of its severity. CNN is able to find out features of disaster appropriately and will helps to differentiate between damaged and un-damaged regions. [ABSTRACT FROM AUTHOR]
- Published
- 2023
- Full Text
- View/download PDF