Browse Prior Art Database

Method and System for Video Surveillance Data Storage Optimization Disclosure Number: IPCOM000248765D
Publication Date: 2017-Jan-06
Document File: 4 page(s) / 126K

Publishing Venue

The Prior Art Database


Disclosed are method and system for video surveillance data storage optimization. The solution includes analysis software that ranks video segments based on the level of severity of any change from the norm, accordingly applies storage dimensions to key frames, and then stores the video frames in the same video file on different layers depending on the storage dimensions.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 44% of the total text.


Method and System for Video Surveillance Data Storage Optimization

To ensure the security of a defined area, video surveillance system capture images of the designated area at all times (e.g., 24 hours per day, every day). Storing video files and the associated data, especially in high resolution, consumes a tremendous amount of server space. However, not storing all data or compromising data quality can pose a security threat.

Video surveillance users need a method and system for storing video files and data that optimize storage space while ensuring the availability of good data quality for security analysis.

The novel contribution is a method and system for video surveillance data storage optimization. The solution includes video storage system and analysis software.

A video surveillance storage server analyzes every video file and accordingly identifies any suspicious event, movement, reorientation of the objects, etc. If no such event is detected, then the video analysis software identifies the key frames within each dynamic time gap and accordingly creates a time-lapse video for those sections. If the system detects any suspicious event, movement etc., then the software keeps that entire video portion.

The video analysis software ranks every event based on severity and accordingly calculates the dynamic time lapse between any pair of frames. For the event of the highest severity, the system stores the entire video section with the same original quality and image dimension. For sections of video that contain no or a low level of severity, the system stores the highest time-lapse video with the same image quality and dimension.

The video storage system identifies various ranges of time-lapse video sections, and accordingly creates another video layer with a key frame dimension of those sections with same video image quality as the original video feed. This reduces overall space consumption of the video files in the server. Therefore, when any security officer wants to view the video surveillance feed, that user can view the continuous video, but the lower priority sections have key frame dimensions.

The video analysis system identifies the most essential image objects (e.g., facial image, etc.) from the time-lapse video section and accordingly overlays those sections on the video screen. This allows the user to map those image objects with lower frame dimension video contents.

To implement the method and system for video surveillance data storage optimization in a preferred embodiment:

1. The administrator of video surveillance can define various rules to identify any event and the severity of an event. For example, if any two adjacent frames are duplicate or almost static for a segment of time, then that instance


is considered less severe. At the same time, if any new object appears, then it might be considered as new essential frame. (Event detection is a known art. Users can also define various event detections in the video file. T...