You have no items in your shopping cart.
ABSTRACT
Video surveillance is crucial in today's society. In public venues like markets, malls, and hospitals, video surveillance systems are widely employed. Increasing public safety through banks, streets, educational institutions, city administration offices, and smart cities, assets, and lives. The primary goal of these security application systems is to make quick and accurate detection of anomalies in videos. Human behavior is erratic, and it can be quite challenging to determine what it is normal or suspicious. This project presents an automated way of detecting anomalies in surveillance videos using deep learning methods. The deep learning model was trained using a spatiotemporal autoencoder architecture, using normal videos for crowded scenes. An anomaly is easier to detect when an abnormal crowd scene is tested with the model. For this solution to be tested, the model was deployed to a web server and can be accessed via a public Application Programming Interface (API). Django web development framework (written in Python programming language) was used with the Django Rest Framework library. This library is very useful for creating RESTful Web APIs that will make our model accessible through HTTP Post requests.
