Distributed-Intelligence Video Surveillance for Moving-Incident Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video surveillance systems lack efficient distribution of intelligence between video cameras, leading to suboptimal incident detection and tracking, especially when incidents move out of one camera's field of view and into another.
Innovation Solution
A video surveillance system with distributed intelligence, utilizing a lead video camera and child video cameras that communicate using different communication protocols, where child cameras generate and transmit metadata to the lead camera, which then relays this information to other cameras and a remote site for recording.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If a central server architecture is used to collect and analyze video data from all cameras, then centralized control and data storage are achieved, but system response time increases and real-time coordination between cameras deteriorates
Solution Approach 1:
The system divides cameras into hierarchical groups with lead cameras and child cameras. Each group operates semi-autonomously, with lead cameras making local decisions about event detection and camera coordination. This segmentation reduces the communication overhead to central servers and enables faster local responses while maintaining centralized architecture for overall system control.
2Reliability
If video data from all cameras is continuously transmitted to a central server for analysis, then comprehensive incident detection is achieved, but network bandwidth consumption increases and system complexity increases
Solution Approach 1:
The system extracts only relevant metadata (event type, location, timestamp, detected objects) from video streams and transmits this compact data structure between cameras and to servers, rather than transmitting full video frames. This extraction approach maintains comprehensive incident detection capability while dramatically reducing network bandwidth consumption and system complexity.
Solution Approach 2:
Cameras perform preliminary event detection and filtering locally before transmitting data to the server. Child cameras detect events in their fields of view and only transmit metadata about detected incidents to lead cameras and the server, rather than continuously streaming all video data. This preliminary action reduces unnecessary data transmission and system complexity.
3Device complexity
If each camera operates independently without inter-camera communication, then device simplicity is maintained, but incident tracking across multiple fields of view deteriorates
Solution Approach 1:
Lead cameras serve as intermediaries between child cameras and the central server. Child cameras communicate event metadata to their lead camera, which then coordinates with other lead cameras and transmits consolidated data to the server. This intermediary structure enables incident tracking across multiple fields of view while maintaining relative system simplicity through hierarchical communication.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An event of interest may be tracked using a distributed intelligence video surveillance system. The distributed intelligence video surveillance system includes a lead video camera and a plurality of child video cameras that are configured to communicate with the lead video camera. One of the plurality of child video cameras may receive a video image and perform analytics and event generation for any events of interest found within the video image. The child video camera extracts metadata for any found events of interest and communicates the extracted metadata to the lead video camera. The lead video camera communicates the extracted metadata to others of the plurality of child video cameras so that the others of the plurality of child video cameras can be alerted to track the found events of interest.