Surveillance Video Storage Reduction via Key Frame Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Surveillance video data storage requires significant space, especially with multiple cameras generating large amounts of data, leading to high storage needs that are costly and inefficient.
Innovation Solution
A computing device implements a storage algorithm that discards a subset of video frames and uses machine learning, such as extreme learning machines, to regenerate and reconstruct the video segments upon request, reducing storage needs by half without compromising video quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all video frames from multiple cameras are stored, then video quality and completeness are maintained, but storage space requirements increase significantly
Solution Approach 1:
The patent extracts and stores only key frames (I-frames) from the video stream, discarding redundant P-frames and B-frames. This selective extraction maintains essential video information while significantly reducing storage requirements by eliminating duplicate or less important frame data.
Solution Approach 2:
The patent creates compressed representations of video data by storing key frames that can serve as references for reconstructing other frames. These key frames act as copies that capture the essential information needed to recreate the full video sequence through frame interpolation and prediction algorithms.
2Quantity of substance
If a subset of video frames is stored to reduce storage space, then storage efficiency improves, but video quality and completeness deteriorate
Solution Approach 1:
The patent performs preliminary encoding and selection of key frames during the video capture phase, organizing and compressing the data before full storage is required. This preliminary processing identifies and preserves the most important frames while discarding redundant data, ensuring storage efficiency is established upfront rather than requiring later compression that might degrade quality.
Solution Approach 2:
The patent changes the representation parameters of video data by converting full video sequences into compressed key frame representations. This parameter transformation allows the same visual information to be stored more efficiently by changing from storing every frame to storing only essential reference frames with metadata for reconstruction.
3Adaptability or versatility
If more cameras are deployed for enhanced surveillance coverage, then monitoring capability improves, but storage requirements increase proportionally
Solution Approach 1:
The patent implements a universal compression algorithm that processes video data from multiple cameras using the same key frame extraction methodology. This multi-functional approach allows the system to handle any number of camera inputs with consistent storage efficiency, enabling scalable surveillance deployment without proportional storage increases.
Solution Approach 2:
The patent segments video data from multiple cameras into individual streams that can be independently compressed and stored. Each camera's video feed is processed separately to extract key frames, allowing for efficient storage management where each camera contributes only its essential frames to the overall storage pool rather than requiring unified storage of all camera data.
4Duration of action of stationary object
If video data is retained for longer periods for security analysis, then investigative capability improves, but storage costs and resource usage increase
Solution Approach 1:
The patent implements periodic retention of compressed video data where only key frames are stored at regular intervals rather than continuous video streams. This periodic sampling approach allows for extended retention periods because the storage burden increases linearly with time rather than exponentially, as each time period adds only a manageable set of key frames rather than complete video sequences.
Data Source
AI summary
In some examples, a computing device may receive a request for a video segment captured by a particular camera, where the request specifies a date, a start time, and a length of the video segment. The computing device may identify stored data associated with the video segment in a storage device based on the date, the start time, the length, and an identifier associated with the particular camera and retrieve the stored data from the storage device. The computing device may determine that the stored data includes a subset of the video frames that were sent from the particular camera and excludes a remainder of the video frames and regenerate the remainder of the video frames based on the stored data to create regenerated data. The computing device may reconstruct the reconstructed video segment by merging the stored data with the regenerated data and provide the reconstructed video segment.