Video Event Location via Contour-Coded Compressed Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large surveillance systems face challenges in efficiently searching through vast video data to identify specific events due to the voluminous metadata representation, leading to impractical network traffic and database size issues, making it difficult for human operators to manually review or use automatic video object tracking effectively.
Innovation Solution
A method involving a spatial representation tool that detects objects, converts trajectory data into contour-coded compressed images, generates searchable codes, and associates them with databases, allowing for efficient querying and retrieval of event locations within video image data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual review of video tapes/files is used to identify events, then complete coverage of the entire monitoring period is achieved, but the time and effort required increases dramatically due to hundreds of hours of video review
Solution Approach 1:
The patent creates a simplified copy or representation of the video data through metadata generation. Instead of reviewing actual video tapes, operators work with generated metadata that captures essential event information, allowing rapid search and identification without time-consuming manual video review while maintaining reliable event detection
Solution Approach 2:
The patent introduces metadata as an intermediary between the raw video data and the operator. This metadata layer automatically generated by the system serves as a mediator that summarizes video content, enabling operators to identify events through metadata queries rather than directly reviewing hundreds of hours of video footage
2Productivity
If automatic video object tracking with standard relational database is used, then event search capability is improved, but the network traffic volume and database table size become impractically large
Solution Approach 1:
The patent extracts only the essential and relevant information from video data to create metadata, rather than storing or processing all video data. This extraction approach generates minimal metadata that contains only the necessary event information, reducing network traffic and database size while maintaining effective search capability
Solution Approach 2:
The patent segments video data processing into distinct stages: video capture, metadata generation, and event search. By separating these functions and processing only essential information at each stage, the system avoids generating voluminous metadata while maintaining effective event search capability through targeted metadata queries
Data Source
AI summary
An invention for identifying a spatial location of an event within video image data is provided. Disclosed are embodiments for detecting an object and obtaining trajectory data of a trajectory of the object within the video image data from a sensor device; converting the trajectory data into a contour-coded compressed image; generating, based on the trajectory data, a searchable code that contains a set of locations traversed by the trajectory of the object within the video image; associating the searchable code with the contour-coded compressed image in a database; and returning, in response to a query having a selected location that corresponds a location of the set of locations in the searchable code, an image of the trajectory data corresponding to the object based on the contour-coded compressed image in the database.