Comparative Analysis on Classification of Coastal Images Using Neural Network

Comparative Analysis on Classification of Coastal Images Using Neural Network

Suvvari Sravani^1, Kalva Jaswanth Babu², Pasumarthi Sowmya³ ,Palempati Harsha Deep⁴,

1,2,3,4, Department of Computer Science and Engineering, GST, GITAM University, Visakhapatnam, AP, India.

Abstract:

Despite recent advancements, classifying the coast remains a difficult problem in quantitative `geomorphological and environmental studies. Years passing It is being a serious issue when land is destroyed or lost due to coastal erosion, especially in a nation like India where population growth is exploding .This had motivated researchers to make a thorough study of the Coastal regions. Years later researchers of Machine learning implemented algorithms for coastal image classification. Our paper is a comparative study and classification of coastal images using neural network by implementing different algorithm and architectures which helps you to give best accuracy in classifying these coastal images. The algorithm we have chosen are k-means algorithm and we have used are CNN (convolutional neural network) deep learning algorithm. In this study we have trained samples of coastal images through these algorithms and tried to increase the accuracy by using the best algorithm between these and used GAN model for data Augmentation that makes accuracy much better. This also helps in predicting which areas are danger prone and which are not.