Video Sequence Analysis Using Spatio-Temporal Activity Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video surveillance systems face challenges in efficiently managing multiple video sequences in real-time due to limited calculation capacity and bandwidth constraints, particularly when simultaneous alarms from multiple cameras cause network congestion, requiring users to manually adjust video quality and leading to suboptimal video transmission.
Innovation Solution
A method and device that analyze video sequences by separating temporal and spatial activity information, transmitting only necessary data, and dynamically adjusting the quality of video streams based on detected movement, using a hybrid encoding method that distributes information across spatio-temporal hierarchical levels, allowing the receiving apparatus to manage alarm detections and optimize data transmission without user intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple video sequences are transmitted simultaneously from multiple cameras, then the user can monitor all surveillance points, but network congestion occurs and bandwidth is exceeded
Solution Approach 1:
The patent segments video data into spatio-temporal hierarchical levels, separating essential temporal activity information from detailed spatial information. Only necessary data is transmitted over the network, reducing bandwidth consumption and avoiding congestion while maintaining monitoring reliability.
Solution Approach 2:
The patent extracts and transmits only the most critical temporal activity information from video sequences, leaving detailed spatial data at local cameras. This extraction approach reduces network traffic significantly while preserving the ability to detect and respond to important events.
2Productivity
If high calculation capacity is used to decode multiple video sequences in real-time, then the user can view all videos simultaneously, but the cost and complexity of the system increases
Solution Approach 1:
The patent extracts computational tasks from the central computer and relocates them to distributed cameras. Each camera independently analyzes temporal activity and makes local decisions about what data to transmit, eliminating the need for high-powered central processing while maintaining real-time monitoring capability.
Solution Approach 2:
The system enables cameras to autonomously analyze video sequences, detect temporal activity patterns, and self-determine what data needs transmission. This self-service approach distributes computational workload and eliminates dependency on high-capacity central computers.
3Manufacturing precision
If manual quality adjustment is implemented, then the user can optimize video quality, but user intervention is required and response time increases
Solution Approach 1:
The patent implements automatic quality adjustment where the system monitors temporal activity and autonomously determines optimal transmission parameters. When significant events are detected, the system automatically adjusts video quality and transmission priority without requiring user intervention, maintaining both quality and responsiveness.
Solution Approach 2:
The system continuously monitors temporal activity patterns and uses this feedback to dynamically adjust video transmission quality. This closed-loop control enables automatic optimization of video parameters based on actual monitoring needs, eliminating manual adjustment delays.
4Loss of information
If all video data is transmitted to the central computer, then complete video information is available, but bandwidth constraints are violated
Solution Approach 1:
The patent segments video information into hierarchical levels of spatio-temporal detail. Only the essential temporal activity data is transmitted over the network, while detailed spatial information remains locally stored. This segmentation maintains information completeness for detection purposes while dramatically reducing data volume for transmission.
Solution Approach 2:
The system maintains high local quality for spatial detail at each camera while transmitting only necessary temporal quality information to the central computer. This local quality approach preserves complete video information where needed while minimizing network data requirements.
Data Source
AI summary
The invention relates to a method of analyzing at least one video sequence present on at least one sending apparatus, in a communication network also comprising at least one receiving apparatus, wherein, the video sequence being compressed with an encoding method which generates a bitstream comprising information representing the temporal activity of the images of the video sequence and information representing the spatial activity of the images of that video sequence, and the temporal and spatial information being distributed between one or more spatio-temporal hierarchical levels, the method comprises the following steps performed at the receiving apparatus:obtaining information representing the temporal activity of the video sequence present on the sending apparatus,analyzing the temporal information obtained,as a function of the result of the analysis, deciding whether the generation of a request destined for the sending apparatus for the obtainment of information representing the spatial activity of the video sequence is necessary.