Vehicle Occupant Sitting Position Assessment With 2D Landmarks

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing driver monitoring systems lack a cost-effective and efficient method to assess the sitting position of a vehicle occupant, particularly the driver, to detect potentially dangerous behaviors such as leaning forward or backward, which can be indicative of drowsiness or eyesight issues.

Innovation Solution

A method using a 2D image analysis with pre-established anatomical landmarks and machine learning algorithms to determine the sitting position of a vehicle occupant, employing off-the-shelf cameras and deep learning for pose estimation, and setting a threshold value during calibration to assess leaning behaviors.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If 3D imaging systems with time-of-flight measurements are used to assess sitting position, then measurement precision is improved, but device complexity and cost increase

Engineering Contradiction:
Improvesitting position assessment accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent uses a 2D camera to capture images and creates a virtual copy of the 3D sitting position assessment through image processing. By detecting anatomical landmarks in 2D images and computing their distances, the system replicates the functionality of expensive 3D imaging systems without requiring time-of-flight measurements or complex 3D sensors.

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent replaces complex mechanical/optical 3D measurement systems with a simpler 2D image processing system. Instead of using time-of-flight cameras or structured light systems, the invention uses standard 2D cameras combined with machine learning algorithms to detect landmarks and infer 3D position information from 2D coordinates.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If 3D imaging systems with time-of-flight measurements are used to assess sitting position, then measurement precision is improved, but cost increases

Engineering Contradiction:
Improvesitting position assessment accuracyVSAvoidsystem cost
Core Design Contradiction:
Measurement precisionVSEase of manufacture

Solution Approach 1:

The patent employs inexpensive 2D cameras that are widely available and cost-effective compared to specialized 3D imaging systems. By using off-the-shelf camera modules and processing software rather than expensive time-of-flight sensors, the system achieves similar functional outcomes at a fraction of the cost.

Inventive Principle:
Principle #27Cheap short-living objects (Disposable)

Solution Approach 2:

The system creates a cost-effective virtual replica of 3D assessment capabilities through 2D image analysis, eliminating the need for expensive proprietary 3D hardware while maintaining the essential functionality of sitting position monitoring.

Inventive Principle:
Principle #26Copying

3Measurement precision

If 3D imaging systems with time-of-flight measurements are used to assess sitting position, then measurement precision is improved, but computational time increases

Engineering Contradiction:
Improvesitting position assessment accuracyVSAvoidcomputation time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent uses 2D image copies rather than full 3D data processing. By analyzing 2D coordinates of anatomical landmarks in captured images and computing distances between them, the system avoids the computationally intensive processes required for true 3D reconstruction while still deriving accurate sitting position information.

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent extracts only the essential information needed for sitting position assessment from images - specifically the 2D coordinates of key anatomical landmarks. By focusing on extracting and processing only these critical data points rather than processing complete 3D point clouds or volumetric data, the system significantly reduces computational overhead.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentEP4632695A1A method for an assessment of a sitting position of an occupant inside a vehicle and an associated monitoring system
Publication Date: 2025.10.15 VALEO COMFORT & DRIVING ASSISTANCE
  • EP4632695A1 patent drawingFigure 1~3
  • EP4632695A1 patent drawingFigure 4~5
  • EP4632695A1 patent drawing

AI summary

A method for an assessment of a sitting position of an occupant inside a vehicle, the method being characterized by the following steps: a. capturing (S21) an image of the occupant, b. analyzing (S22) the image of the occupant to provide 2D coordinates in the image of pre-established anatomical landmarks and parameters representative of a pose of the occupant, c. computing (S23) distances using said 2D coordinates of said pre-established anatomical landmarks, d. normalizing (S24) the computed distances using the parameters representative of a pose of the occupant, e. computing (S25) a mean anatomical distance using the normalized distances, f. comparing (S26) the mean anatomical distance to a threshold value to assess the sitting position of said occupant. A corresponding monitoring system is also described.