AI-Based Selective Video Compression for Storage Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge is to maintain high video quality for important sequences in video streams while conserving storage space, as existing methods of compressing video streams to reduce storage requirements often result in reduced video quality, making it difficult to identify important details.
Innovation Solution
A computing device uses a machine-learning model to detect objects-of-interest in video streams and selectively vary audio or video compression based on these classifications, preserving high video quality for sequences-of-interest while reducing storage requirements for less critical frames.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If the entire video stream is compressed to reduce storage size, then storage space is conserved, but video quality is reduced making it harder to view important details
Solution Approach 1:
The video stream is divided into individual frames, and each frame is evaluated separately by the machine learning model to determine if it contains an object of interest. This allows selective compression at the frame level rather than uniform compression of the entire video stream.
Solution Approach 2:
Different compression quality levels are applied to different portions of the video stream based on their importance. Frames containing objects of interest are preserved at high quality, while frames without objects of interest are compressed more aggressively, creating locally optimized quality distribution.
2Manufacturing precision
If high-quality video is maintained for all frames, then video quality is preserved, but storage space requirements increase significantly
Solution Approach 1:
The compression parameter (quality level) is dynamically changed based on the content of each frame. The machine learning model detects objects of interest and adjusts the compression parameter accordingly, using higher quality settings for important frames and lower quality settings for less important frames.
Solution Approach 2:
The compression quality is made dynamic rather than static. The system continuously analyzes incoming frames and adjusts compression levels in real-time based on detected object interest, allowing the video stream to adapt its quality profile to match actual content importance.
Data Source
AI summary
In one embodiment, a computing device receives, from one or more cameras, a video stream comprising multiple frames, where the video stream is received at a first quality. The computing device analyzes, using a machine-learning model, images in the frames, where the machine-learning model has been trained to detect one or more objects-of-interest in the images. The computing device identifies a sequence-of-interest including consecutive frames of the video stream, where at least one object-of-interest was detected in at least one of the consecutive frames. The computing device generates a video package including the sequence-of-interest.