Application of deep learning technique in UAV’s search and rescue operations

Abstract

This paper is concerned with the application of Deep Learning techniques for analyzing image data for search and rescue operations of Unmanned Aerial Vehicle (UAV). It uses Keras and its Tensorflow backend to model a deep Convolutional Neural Network (CNN) Learning technique and train the model with MNIST digits dataset to predict the hand- written word from the image data received from the ground level. The paper explains the stages involved in the implementation of LeNet method of Deep Learning techniques for developing a classifier for long distance recognition of handwritten words.