Sensor-Image Correlation for Worker Identification in Similar Attire
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for identifying individuals in moving images at manufacturing sites, such as those using camera recordings and sensor data, face challenges due to similar work attire and unclear facial recognition, and accuracy issues arise from varying sensor positions.
Innovation Solution
An information processing apparatus that acquires image data and sensor data to generate part movement information and sensor operation information, calculating correlation values to accurately match individuals in images with sensor data without relying on sensor position, using a system with image data acquisition, processing, and identification units.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If face analysis and clothes analysis are used to identify workers in moving images, then worker identification can be performed, but identification accuracy decreases when workers wear similar work clothes and helmets with unclear faces
Solution Approach 1:
The patent introduces sensor data (acceleration, angular velocity, magnetometer values) as an intermediary to bridge the gap between visual appearance and worker identity. Instead of relying solely on visual features that become indistinguishable when workers wear similar attire, the system uses sensor data as a mediator to capture unique movement patterns that identify each worker, thereby resolving the contradiction between identification accuracy and applicability to workers in similar clothing
Solution Approach 2:
The patent replaces the mechanical/visual analysis system (face and clothes analysis) with a sensor-based measurement system. By substituting visual identification methods with sensor data analysis, the system overcomes the limitation of similar work attire and helmets, maintaining high identification accuracy across diverse working conditions where visual features are insufficient
2Measurement precision
If sensor data from mobile terminals is used to identify persons by matching walking pitch and phase, then person identification can be performed, but matching accuracy decreases because sensor position varies depending on where the mobile terminal is worn
Solution Approach 1:
The patent segments the sensor data into multiple independent components (acceleration data, angular velocity data, magnetometer data) and processes each segment separately. By dividing the complex sensor information into distinct segments, the system can apply specific processing methods to each type of data, reducing the impact of position variability on overall matching accuracy
Solution Approach 2:
The patent transitions from analyzing single-dimension data (such as only acceleration) to multi-dimensional sensor data by incorporating angular velocity and magnetometer measurements. This dimensional expansion creates a more robust identification system that is less sensitive to sensor position changes, as the combined multi-dimensional data provides redundant information that compensates for position variability
Data Source
AI summary
According to one embodiment, an information processing apparatus includes processing circuitry. The processing circuitry acquires image data on a predetermined photographed area, generates part movement information on a plurality of parts of an object included in the image data, based on the image data, acquires sensor data from the object, generates sensor operation information on an operation of a sensor, based on the sensor data, and specifies a relationship between the object included in the image data and the sensor, based on the part movement information and the sensor operation information.