Machine learning based Smart Surveillance System

In modern cities video surveillance becomes an important part for protection and security. In modern cities for security and protection smart cameras equipped with intelligent analysis of video can monitor events and give alert system by capturing abnormal activities and events. Now a day's there are many cameras have been equipped for surveillance in all over world. So we are going to implementation focus on video surveillance by giving video contents containing early fire events detection, abnormal activities and smart parking system and crowd estimation. We are going to overcome existing drawbacks of post investigation techniques of video surveillance systems by providing alert generation system. Our work is based on machine learning techniques for video analysis with better performance and event detection with advantages of alert generation.

1. Introduction

The video surveillance and monitoring has become the richest source of security and investigation.in this project we would like to introduce modules like fire detection, crowd detection, smart parking, and abnormal activity detection. The current world is completely under CCTV or video surveillance systems. The video recorded is used to find out robbery investigation, crime investigation and abnormal activity detection. After event happened these video sequences is used to catch criminals. But problem is that after event happened we are unable to save loss done by that event or accidents. So there is need of such systems which is able to early event detection and alert generation systems. Motive behind proposed work to parsed meaningful structured information from the raw non-structured video.