In-Vehicle Skeleton Tracking for Passenger Pole-Hold Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In the riding bus business, there is a need to determine whether a standing passenger is holding onto a pole or similar, which becomes more critical with the transition from manual to autonomous driving. Existing systems require sensors to be installed at multiple handrail or strap locations, which is impractical.
Innovation Solution
An information processing apparatus that analyzes in-vehicle video footage to acquire skeleton information of passengers and determine if they are holding a pole, without the need for sensors at each potential holding point. This system includes a skeleton information acquisition unit, a holding determination unit, and an alert output unit.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If sensors such as contact sensors are installed in handrails and straps to detect whether a passenger is holding on, then the reliability of passenger safety monitoring is improved, but the device complexity and installation cost increase significantly
Solution Approach 1:
The patent replaces the mechanical sensor-based detection system with a vision-based system using cameras and image processing. Instead of installing contact sensors in handrails and straps, the system uses cameras to capture images and algorithms to determine whether passengers are holding on, thereby eliminating the need for complex sensor installations while maintaining monitoring reliability
Solution Approach 2:
The patent creates a virtual model of the passenger's state by analyzing images. It generates skeleton information and determines holding status through image processing rather than direct physical sensing, effectively copying the detection function from the physical sensor domain to the digital image processing domain
2Measurement precision
If multiple sensors are installed at various handrail and strap locations to cover all possible holding points, then the measurement precision of passenger holding status is improved, but the ease of manufacture and installation deteriorates
Solution Approach 1:
The patent makes the camera system universal by positioning it to capture the entire vehicle interior, allowing a single camera to monitor multiple handrails and straps simultaneously. The image processing algorithm universally analyzes all potential holding points within the field of view, eliminating the need for location-specific sensors at each handrail or strap
Solution Approach 2:
The patent transitions from one-dimensional point-based sensor detection to two-dimensional area-based image analysis. By capturing the entire vehicle interior with a camera, the system can detect passenger holding status across multiple locations simultaneously through spatial analysis of the image data
3Reliability
If contact sensors are installed in all handrails and straps that can be held by passengers, then the completeness of passenger safety monitoring is improved, but the loss of substance and installation resources increase
Solution Approach 1:
The patent replaces physical contact sensors with optical cameras and digital image processing. This substitution eliminates the need for numerous physical sensors while achieving comprehensive monitoring coverage through visual analysis of the entire vehicle interior
Solution Approach 2:
The patent creates a digital representation of the vehicle interior and passenger states through image capture and processing. This virtual model allows comprehensive monitoring of all handrails and straps without requiring physical sensors at each location, thereby conserving material resources
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present disclosure enables determination of whether a passenger is holding a pole without installing a sensor, such as a contact sensor, at each of a plurality of places where a passenger may hold. A skeleton information acquisition means (51) analyzes an in-vehicle video acquired from a camera (31) that captures an image of a vehicle interior of a moving object (30), and acquires skeleton information of each of one or more passengers boarding on the moving object (30). A holding determination means (52) determines whether the passenger is holding a pole, based on skeleton information of the passenger and a position of the pole in the moving object (30). An alert output means (53) outputs an alert when it is determined that the passenger is not holding the pole.