• Apoorva Kamath
• Brunda S
• K L Westbenson
• Shubham Sharma
• Niraj Sinam

In this modern era, surveillance cameras are omnipresent. Be it home, office, malls, highway, CCTV cameras are found everywhere monitoring the day to day activities. Surveillance is an important task. They aid in detecting any fishy activity and act as valuable proof. However, the amount of video footage recorded is enormous. Most of the times the useful information contained in it is very less. It becomes a very difficult task to go through so many hours of footage to find the useful content. Storing the videos also takes a huge amount of space. Thus, a need for surveillance solutions is recognized that can condense such long footage into short videos while still retaining important events. Our model can summarize hours of footage shot by CCTV cameras into a short clip that shows all interesting events simultaneously. All moving objects that are present in the video are identified and tracked by our program. These events are then overlaid on a single clip, while also displaying the timestamps for each event, thus letting the user perform surveillance for multiple events. Our model uses KNN classifier for extracting and subtracting background from videos and a custom Object Tracking algorithm to detect moving objects, which is then overlapped on extracted background.

Video Summarization, Retain important events, KNN Classifier, Object Tracking

"Summarization of CCTV Videos using KNN Classifier", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.8, Issue 6, page no.d507-d510, June-2021, Available :http://www.jetir.org/papers/JETIR2106474.pdf

"Summarization of CCTV Videos using KNN Classifier", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.8, Issue 6, page no. ppd507-d510, June-2021, Available at : http://www.jetir.org/papers/JETIR2106474.pdf