Video Data Classification Using Sensor-Driven Activity Codes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video surveillance systems require extensive manual review of large quantities of video data to locate specific events, such as unauthorized activity, due to the lack of efficient classification methods.
Innovation Solution
A method and apparatus that classify video data using sensor data to define classification codes for programme elements within the video, allowing for easier navigation and storage of video streams by identifying and storing only relevant activity levels, using sensors like motion, sound, or pressure sensors to generate classification codes for video clips.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If video data is captured continuously using traditional video cassettes, then complete video records are obtained, but large quantities of video data must be reviewed manually requiring random fast-forward and rewind operations
Solution Approach 1:
The video data is segmented into discrete programme elements (individual video clips) that are independently classified and tagged with classification codes based on sensor data. This segmentation allows the system to process and review only specific segments rather than the entire continuous video stream, reducing manual review time while maintaining complete records.
Solution Approach 2:
Classification codes are assigned to video programme elements in advance during the capture process, before manual review is needed. Sensor data is processed preliminarily to determine activity levels, and classification codes are applied automatically, so that when review is required, the data is already organized and indexed, eliminating the need for random fast-forward and rewind operations.
2Ease of operation
If classification codes are applied to video programme elements using sensor data, then video data becomes easier to navigate and store, but additional processing steps and sensor integration are required
Solution Approach 1:
The system uses multi-functional sensors that can detect various types of activity (motion, sound, pressure) and generate classification codes applicable to different video scenarios. The same classification framework handles different activity types uniformly, simplifying the interface for users while consolidating multiple detection functions into a single manageable system.
Solution Approach 2:
Sensor data serves as an intermediary between the physical monitoring environment and the video data storage system. The sensors detect activity conditions and translate them into classification codes that bridge the gap between raw sensor inputs and video programme elements, enabling automatic classification without requiring direct complex integration between all system components.
3Quantity of substance
If video data representing low activity periods is deleted to optimize storage, then storage space is saved, but potential loss of relevant information may occur
Solution Approach 1:
The system changes the parameter of data retention based on activity level thresholds. Video programme elements are evaluated against predetermined activity criteria, and classification codes determine whether data should be retained or deleted. This parameter-based approach allows automatic optimization of storage space by removing low-value data while maintaining the flexibility to retain important information when activity thresholds are met.
Data Source
AI summary
A method of classifying video data representing activity within a space to be monitored. A method comprises storing video data obtained from a camera configured to monitor the space. Sensor data indicative of a condition occurring within the space is obtained, and a plurality of program elements are defined within the video data. Each program element has an associated classification code, and each classification code is selected using the sensor data.