Image Analysis for Hidden-State Detection of Overlapping Persons
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image analysis systems struggle to accurately capture and analyze images of monitored spaces due to positional relationships with cameras, leading to incomplete or incorrect information acquisition of individuals when multiple persons overlap, hindering the detection of skeleton points and occupant information.
Innovation Solution
An image analysis apparatus equipped with a hidden state detection unit that employs movement prediction, overlap ratio calculation, number-of-persons difference determination, and camera proximity position detection to identify and correct for hidden states where individuals are overlapped or obstructed in the image, ensuring accurate detection of skeleton points and occupant information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single camera is used to capture images of the monitored space, then the device complexity is reduced, but the measurement precision of person information deteriorates due to overlapping persons in the captured image
Solution Approach 1:
The system performs preliminary detection of hidden states by analyzing spatial relationships and movement patterns before final person information extraction. The hidden state detection unit identifies overlapping persons in advance, allowing the system to flag or correct measurement issues before they compromise the final analysis accuracy.
Solution Approach 2:
The system implements feedback by using detection results to improve subsequent measurements. When overlapping persons are detected, the system adjusts its analysis approach, re-examines the captured image with modified parameters, or flags the measurement for manual review, thereby continuously improving measurement precision based on detected conditions.
2Measurement precision
If multiple cameras are deployed to eliminate hidden states, then the measurement precision of person information is improved, but the device complexity increases
Solution Approach 1:
Rather than deploying multiple cameras, the system performs preliminary analysis of single-camera images to detect hidden states. By identifying overlapping persons and occlusion conditions in advance, the system can compensate for single-camera limitations through algorithmic corrections and alternative detection methods.
Solution Approach 2:
The patent replaces the mechanical solution of adding more cameras with an information-processing solution. The hidden state detection unit and analysis algorithms substitute for additional physical sensors, using computational methods to extract person information even when visual data is partially occluded or overlapping.
3Adaptability or versatility
If the camera is positioned to capture the entire monitored space, then the adaptability of the monitoring system is improved, but the measurement precision of individual persons deteriorates due to overlapping
Solution Approach 1:
The system segments the monitoring task by separating hidden state detection from final person information extraction. The hidden state detection unit specifically identifies overlapping conditions, allowing the main analysis unit to apply specialized processing techniques for each detected segment or region, thereby maintaining both wide coverage and individual accuracy.
Solution Approach 2:
The system performs preliminary segmentation of the monitored space by detecting hidden states in different regions. This allows subsequent analysis to be tailored to each region's specific conditions, applying appropriate measurement techniques that account for local overlapping or occlusion patterns while maintaining overall space coverage.
Data Source
AI summary
An image analysis apparatus includes a person information acquisition unit configured to analyze a captured image of a monitored space imaged by a camera so as to acquire information of a person imaged in the captured image, and a hidden state detection unit configured to detect occurrence of a hidden state in which a plurality of the persons are imaged to overlap in the captured image.