Ear Shape Analysis Using Principal Component Weight Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for calculating head-related transfer functions face challenges in accurately estimating ear shapes, leading to misestimation and unrealistic results due to inappropriate image processing, particularly when the head shape of the listener does not match the dummy head used for measurement.
Innovation Solution
An ear shape analysis method using a principal component weight vector to generate an ear shape data set, which identifies an estimated three-dimensional shape of a target ear by applying the weight vector to an ear shape model indicating the relation between ear shape data sets and principal component weight vectors, reducing the probability of misestimation compared to shape deformation methods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a standard head shape is deformed using morphing technique to estimate listener head shape, then the head-related transfer function can be calculated without physical measurement, but the estimated ear shape becomes unrealistic and collapsed when image processing is not carried out appropriately
Solution Approach 1:
The patent transforms the ear shape estimation problem from direct morphing of head shapes to a parameter-based approach using principal component analysis. By representing ear shapes as combinations of principal components derived from multiple sample ears, the system changes the parameter space from continuous shape deformation to discrete component weighting, ensuring realistic ear shapes while enabling easy calculation of head-related transfer functions.
Solution Approach 2:
Instead of directly morphing a standard head shape, the patent creates accurate copies of real ear shapes by combining principal components from multiple sample ears. This copying approach using statistical models ensures that the estimated ear shape faithfully reproduces realistic anatomical features while avoiding the collapse problems of direct morphing techniques.
2Measurement precision
If head-related transfer function is measured directly from the listener's head, then accurate individualized HRTF is obtained, but considerable physical and psychological burdens are imposed on the listener
Solution Approach 1:
The patent enables the listener to receive accurate individualized head-related transfer functions without active participation in the measurement process. By using automatically captured images and principal component analysis, the system performs the estimation task independently, eliminating the need for the listener to undergo time-consuming or uncomfortable physical measurement procedures while still achieving accurate individualized results.
3Extent of automation
If feature points are extracted from head images to estimate head shape, then the process can be automated, but inappropriate image processing leads to misestimation of ear shape
Solution Approach 1:
The patent introduces principal component analysis as an intermediary between image capture and ear shape estimation. Rather than directly extracting feature points from images, the system uses principal components derived from multiple sample ears as an intermediate representation. This intermediary layer filters out processing errors and ensures reliable ear shape estimation even when image processing varies, while maintaining full automation.
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
A computer generates ear shape data by applying a principal component weight vector to an ear shape model indicating a relation between ear shape data sets and principal component weight vectors, each ear shape data set representing a difference between a point group representing a three-dimensional shape of an ear and a point group representing a three-dimensional shape of a reference ear, and each principal component weight vector indicating weights of principal components of the corresponding ear shape data set. From the ear shape data generated, the computer identifies an estimated three-dimensional shape of a target ear corresponding to a target ear image represented by image data.