Intelligent Video Surveillance Systems Using Deep Learning Methods

Abstract

An expeditious growth of data size in real-time environments leads to recent commercialization of deep learning which has propelled various technical advancements in many fields like big data analytics, computer vision, sentiment analysis, health care, agricultural, Internet of Things, machine translation and recommender systems. Deep learning evolved from a branch of machine learning intern from artificial intelligence; primarily, deep learning has two aspects. The aim of this chapter is to propose custom deep learning algorithms which are used in agricultural image processing and techniques used for natural language processing (NLP) for the development of voice assistant and dialog generation systems. Deep learning algorithms are populated in computer vision for accurate object detection in real-time intelligence surveillance systems. This chapter presents solid deep learning algorithms for image classification and object detection. The objective of this chapter is to present customized lightweight deep convolutional neural network (LW-DCNN) for image classification. The model effectively performs in classifying the working condition of insulators which are used on electric polls. The chapter presents a hybrid application to audio signal and describes the analysis of various feature extraction methods to audio signals.