A Review on Anomalous Events Detection and Recognition

Document Type : Original Article

Authors

1 Computer Science Department, Faculty of Computers and Artificial Intelligence, Helwan University, Cairo, Egypt

2 Computer Science Department, Faculty of Computer and Artificial Intelligence, Helwan University, Cairo, Egypt

Abstract

Anomalous behavior may indicate threats and dangers to others. We can define anomalous event as what drift from what is anticipated, common, or natural action. Which make video surveillance a key to increase public security. The main goal of event detection is to detect the occurrences of events and categorize them into normal or abnormal actions. Such detection requires detecting and tracking objects then recognize what is happening around those tracked objects. Recently, researches depend on one of two techniques: handcrafted features and deep learning Models. The handcrafted features depend on extracting low-level features, its strength depend on choosing the best features, which gives the best results. After the successful of deep learning techniques for image classification, researchers have examined the deep learning techniques ability for detection, which skip the manual step of features extraction and deals directly with the images. This paper presents a survey on both handcrafted and deep learning models for abnormal events detection.

Keywords

Main Subjects

Files

Share

How to cite

Statistics