Reference-Timed Room Audio-Video Processing for Facility Abnormalities
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The high processing load on devices when detecting abnormalities in facilities by analyzing audio and video data acquired from rooms is a significant challenge.
Innovation Solution
A data processing device and method that selectively processes a part of the acquired audio or video data using a reference timing set in advance to generate facility information, reducing the load by identifying specific conditions such as movement or keywords, thereby decreasing the overall computational requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the process of detecting an abnormality in a facility or a sign thereof is performed for all the audio data and the video, then the detection accuracy is improved, but the load on a processing device increases
Solution Approach 1:
The patent segments the continuous audio and video data into specific time periods based on facility operation status. The processing unit identifies and processes only the audio/video data corresponding to periods when the facility is operating, rather than analyzing all recorded data. This segmentation reduces the volume of data requiring complex processing while maintaining the ability to detect abnormalities accurately during relevant operational periods.
Solution Approach 2:
The patent applies partial action by processing only a subset of the acquired audio and video data - specifically, the portions corresponding to facility operation periods. The processing unit selectively analyzes data during identified operational windows, performing sufficient processing to detect abnormalities without the excessive computation required for complete continuous analysis of all recorded data.
2Reliability
If the processing is performed on all acquired data, then the reliability of abnormality detection is improved, but the processing time increases
Solution Approach 1:
The patent performs preliminary action by first identifying the time periods when the facility is operating before conducting abnormality detection processing. The processing unit uses the operation status information to pre-select relevant audio and video data segments, so that the subsequent abnormality detection is performed only on predetermined relevant periods, reducing overall processing time while maintaining detection reliability.
Solution Approach 2:
The patent segments the data processing task into two distinct phases: first identifying operation periods based on facility status, then performing abnormality detection only on data from those identified periods. This segmentation of the processing workflow ensures that computational resources are focused on relevant time windows, maintaining reliable detection while minimizing unnecessary processing time spent on data from non-operational periods.
Data Source
AI summary
A data processing device is a device that detects an abnormality in a facility disposed in a room or a sign thereof and includes an acquisition unit and a processing unit. The acquisition unit acquires processing target data. The processing target data is generated by recording an audio or a video inside the room. The processing unit processes a part of the processing target data (hereinafter, referred to as a processing target part) for generating facility information relating to a facility disposed inside the room. Then, the processing unit detects a reference timing that is a timing satisfying a reference condition set in advance and sets the processing target part using the reference timing.