Prunable Video Encoding via Subview Segmentation and Hierarchical P-Frames
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video surveillance systems face challenges in efficiently managing storage space due to the large size of video data, which requires either lengthy storage or labor-intensive pruning processes, often involving decoding and significant computational load.
Innovation Solution
A method of encoding video streams by dividing each frame into subviews with assigned categories, using a hierarchical P-frame structure, allowing for efficient pruning while maintaining a full view of the scene, and enabling predictable pruning effects by differentiating the importance of scene parts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If video data is stored for long periods to ensure surveillance coverage, then reliability of surveillance is improved, but storage space consumption increases
Solution Approach 1:
The video stream is divided into multiple subviews, each representing a different region or perspective of the surveillance scene. This segmentation allows selective retention of important subviews while pruning less critical ones, thereby reducing overall storage requirements while maintaining essential surveillance coverage.
Solution Approach 2:
Different pruning strategies are applied to different subviews based on their importance. Critical subviews that contain important surveillance information are retained longer, while less important subviews are pruned more aggressively. This local differentiation optimizes the balance between storage efficiency and surveillance reliability.
2Quantity of substance
If straightforward FIFO pruning is applied to reduce storage space, then storage space is freed, but surveillance reliability deteriorates
Solution Approach 1:
Instead of applying uniform FIFO pruning across all video data, the invention applies differentiated retention policies to different subviews. Subviews containing critical surveillance information are marked for longer retention, while less important subviews are pruned first. This ensures that storage space is freed efficiently while maintaining surveillance reliability for important regions.
3Quantity of substance
If video quality is decreased to reduce file size, then storage space consumption is reduced, but video quality deteriorates
Solution Approach 1:
The video is segmented into subviews that can be encoded with different quality levels. Important subviews are maintained at higher quality, while less critical subviews use lower quality encoding. This selective quality differentiation reduces overall storage requirements without significantly impacting the quality of important surveillance data.
4Productivity
If pruning is performed without analyzing actual image data, then computational load is reduced, but pruning effectiveness decreases
Solution Approach 1:
The video stream is pre-encoded with a hierarchical P-frame structure and divided into categorized subviews during the encoding phase. This preliminary organization embeds metadata about subview importance and temporal relationships. During pruning, this pre-established structure allows efficient decision-making without requiring complex real-time analysis of image content, thus maintaining both high pruning efficiency and effectiveness.
5Productivity
If hierarchical P-frame encoding structure is used with category monitoring, then pruning efficiency is improved, but device complexity increases
Solution Approach 1:
The video encoding is segmented into subviews with assigned categories using standard video coding techniques. Each subview is independently encoded with P-frames that reference previous frames, creating a hierarchical structure. This segmentation approach leverages existing encoding standards while adding categorical organization that enables efficient pruning without requiring fundamentally new encoding algorithms.
Solution Approach 2:
The hierarchical P-frame structure serves multiple functions: it provides standard video compression, establishes temporal relationships between frames, and enables efficient pruning operations. By making the encoding structure multi-functional, the invention avoids adding dedicated complexity solely for pruning purposes, as the same structure supports both compression and selective retention.
Data Source
AI summary
A method of encoding a video stream for the provision of prunable video data, comprising: receiving, in an encoder, a video stream in which each image frame of the video stream is divided into a number of subviews, providing each subview with an assigned category based on a content of the particular subview, monitoring the evolution of category assigned for each subview, and encoding the video stream using a hierarchical P-frame encoding structure, in such manner that each change in category for a subview starts a new hierarchical P-frame structure.