Motor Vehicle Object Rating via 3D Image Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing motor vehicle systems for contactless operation of devices, such as gesture control and object detection, lack reliability in accurately determining object states like hand positions and gestures, especially when objects are inclined or irregularly shaped.

Innovation Solution

An object evaluation device using a PMD sensor and digital image processing to create a three-dimensional image, segmented into sectors, determines spherical curvatures of objects like hands, allowing for reliable state evaluation and device operation through sector points and enveloping spheres, enabling precise control without contact.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If gesture control systems are used for contactless operation, then ease of operation is improved, but reliability of accurate detection deteriorates

Engineering Contradiction:
Improvecontactless operationVSAvoiddetection accuracy
Core Design Contradiction:
Ease of operationVSReliability

Solution Approach 1:

The three-dimensional image space is divided into multiple sectors, and sector points are extracted from each sector to represent local geometric features. This segmentation approach enables reliable detection of hand gestures by analyzing curvature characteristics in different spatial regions, resolving the contradiction between contactless operation ease and detection reliability.

Inventive Principle:
Principle #1Segmentation

2Device complexity

If simple object detection methods are used, then device complexity is reduced, but measurement precision deteriorates

Engineering Contradiction:
Improvedetection system complexityVSAvoidobject state detection precision
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

Spherical curvature is calculated from sector points to characterize the geometric shape of detected objects (e.g., hand gestures). This curvature-based measurement provides high precision in determining object states while maintaining relatively simple computational processing, thus resolving the contradiction between device complexity and measurement precision.

Inventive Principle:
Principle #14Spheroidality (Curvature)

3Speed

If traditional 2D image processing is used, then processing speed is improved, but detection accuracy for 3D gestures deteriorates

Engineering Contradiction:
Improveprocessing speedVSAvoidgesture detection accuracy
Core Design Contradiction:
SpeedVSMeasurement precision

Solution Approach 1:

The system processes three-dimensional image data by extracting sector points from 3D space and calculating spherical curvature, rather than using traditional 2D image processing. This dimensional approach maintains processing efficiency while significantly improving gesture detection accuracy by capturing spatial information that 2D methods cannot detect.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

Enables safe and reliable contactless operation of motor vehicle devices by accurately detecting various hand states and gestures, allowing occupants to focus on the road while controlling systems effectively.

Implementation Method 1

The object evaluation device includes, for example, a camera with a PMD sensor (PMD=Photonic Mixing Device), which is based on the principle of the time-of-flight method

Methodology Applied
Scientific EffectTime-of-flight method: Time of Flight

Data Source

PatentEP3221772A1Method for operating an object rating apparatus for a motor vehicle, object rating apparatus for a motor vehicle and motor vehicle having an object rating apparatus
Publication Date: 2017.09.27 AUDI AG

AI summary

The invention relates to a method for operating an object rating apparatus (20) for a motor vehicle (10), in which at least one object state (24, 26) of an object (22) to be rated is optically captured by at least one sensor (12) of the object rating apparatus (20) and a three-dimensional image (28), based on the capture, of the object (22) is segmented, the segmented three-dimensional image (28) being divided into a plurality of sectors (30, 32, 34, 36). To describe the object state (24, 26), at least one sector point (38) from a plurality of physically distributed sector points of the image (28) that describe the object (22) is used from each of the sectors (30, 32, 34, 36). The plurality of sector points is taken as a basis for determining at least one spherical curvature (40) of the object (22), wherein the object state (24, 26) is rated on the basis of the spherical curvature (40) of the object (22) and the rated object state (24, 26) is used as a state signal for operator control of an appliance (48) of the motor vehicle (10).