COMPUTER VISION-BASED TECHNIQUE FOR OBJECTIVE ASSESSMENT OF MATERIAL PROPERTIES IN NON-RIGID OBJECTS

Patent №

US 6,594,381

Granted

2003-07-15

Filed 2001

Owner

SOUTH FLORIDA, UNIVERSITY OF

+2 more

Lab

AI components

2

vision · planning

Assignment

Recorded

Dataset

AIPD

2023_r1 edition

Application

09858064

The present invention provides a system and method for recovering material properties of non-rigid objects. The method includes, for example, the steps of: establishing a plurality of three-dimensional point correspondences of the non-rigid object in an unstressed and stressed state; from the plurality of point correspondences, generating a finite element model of the non-rigid body having initial material properties and generating a finite element strain distribution; detecting abnormal areas of the non-rigid body by comparing finite element strain levels from the strain distribution; and determining at least one material property of the abnormal areas. The system includes, for example, a general or personal computer with, optionally, one or more digital cameras and range finding devices for acquiring and processing three-dimensional correspondence data associated with the non-rigid object being an unstressed and stressed state. Among the plurality of material properties recoverable by the present invention are material elasticity and geometry. An iterative descent method is utilized to refine the finite element model and recover the missing or estimated material properties.

VisionPlanningG06T 7/0012G06T 7/246

AI classification

Planning1.00
Vision0.97
Knowledge representation0.01
AI hardware0.00
Machine learning0.00
Natural language0.00
Evolutionary computation0.00
Speech0.00

Ownership

SOUTH FLORIDA, UNIVERSITY OF

assignment · 139870808

FLORIDA, UNIVERSITY OF SOUTH

correct · 140970896

UNIVERSITY OF SOUTH FLORIDA

correct · 146990893

© 2026 NYSGPT2525 LLC