Understanding the behavior of humans is a very important concern for social communication. Especially in real-time, predicting human activity and behavior has become the most vigorous research area in digital image processing and computer vision. To enhance the security in public and private domains in the field of humancomputer interaction and intelligent video surveillance, human behavior analysis is an important challenge in various applications. There are many basic approaches to analyze human activity, but recently, deep learning approaches have been shown that yield very interesting results in different domains. Human actions and behavior can be observed in the open as well as in sensitive areas, such as airports, banks, bus and train station, colleges, parking areas, etc., and prevent terrorism, theft, accidents, fighting, as well as other abnormal and suspicious activities through visual surveillance. This chapter thus seeks to reflect on methods of human activity recognition. This chapter presents a brief overview on human behavior recognition along with its challenges or issues and applications. Also, we have discussed the framework of recognition of suspicious human activity and various datasets used to train the system. The objective of this chapter is to provide general information about human behavior analysis and recent methods used in this field.
Keywords: Activity recognition, Convolutional neural network, Deep learning, Feature extraction, Human behavior analysis, Object classification, Object segmentation.
