ORIGINAL RESEARCH A review of supervised and unsupervised machine learning techniques for suspicious behavior recognition in intelligent surveillance system Kamal Kant Verma 1 • Brij Mohan Singh 2 • Amit Dixit 3 Received: 29 July 2017 / Accepted: 14 September 2019 Ó Bharati Vidyapeeth’s Institute of Computer Applications and Management 2019 Abstract There is a strong demand of smart vision based surveillance system owing to the increase in crime at a frightening rate at various public places like Banks, Air- port, Shopping malls and its application in human activity recognition ranges from patient fall detection, irregular pattern recognition or Human computer Interaction. As the crime increases at a disturbing rate, public security viola- tions and high cost of security personals have motivated the author to do the strategic survey of existing vision and image processing based techniques in the past literature. The paper begins with discussing the common approach towards suspicious activity detection and recognition fol- lowed by summarizing the supervised and unsupervised machine learning methodologies mainly based on SVM, HMM and ANN classifiers, which were adopted by the researchers previously varying from single human behavior modeling to crowded scenes. Next, this paper discusses system model for human’s normal and abnormal activities recognition along with various feature selectors and detectors used in previous literature. This was followed by conducting a review of benchmark researches which cov- ered a comprehensive state of art methodologies in the related fields, key points owned, feature learning and applications. At last experimental aspects of various papers have been discussed with essential performance matrices like accuracy along with the major issues, common prob- lems, challenges and future scope in the related field. Keywords Abnormal activity Á Intelligent surveillance Á Object detection Á Object tracking and supervised and unsupervised learning 1 Introduction to video surveillance system Over the past few years, crime has substantially increased due to which, there arises the need of intelligent video surveillance system rather than manual surveillance. Manual surveillance system was restricted to small places and it required security guards to monitor any abnormal activity. Manual surveillance is costly and required more efforts to find any suspicious activity. Now-a-days it is not possible to monitor all the activities with the help of security guards. Due to this reason we need an intelligent video surveillance system. Intelligent video surveillance system observes the sensitive areas which are vulnerable to the crime like violence in ATM and banks, theft, fire detection, fight detection etc. [1] As the crime rate is increasing at very high speed so we need an automatic video surveillance system that could detect abnormal behavior, any suspicious activity and could play a vital role in the security because it recognizes the unusual event at any stage in the video sequence. As we know that finding an anomaly in the video is an emerging research problem to maintain surveillance in captured video that consists of detecting, classifying, tracking, and recognizing the behavior of an object. A lot of techniques have been designed by the various & Kamal Kant Verma kkv.verma@gmail.com Brij Mohan Singh bmsingh1981@gmail.com Amit Dixit dixitamit777@gmail.com 1 Uttarakhand Technical University, Dehradun, India 2 Department of CSE, College of Engineering Roorkee, Roorkee, India 3 Department of ECE, Quantum University, Roorkee, India 123 Int. j. inf. tecnol. https://doi.org/10.1007/s41870-019-00364-0