METHODS USING SPECIFIC ATTRIBUTES AND GRAPH GRAMMARS IN GRAPH-BASED TECHNIQUES FOR FEATURE RECOGNITION

Patent №

US 7,042,451

Granted

2006-05-09

Filed 2003

Owner

GEOMETRIC SOFTWARE SOLUTIONS CO., LIMITED

+1 more

Lab

AI components

3

ml · vision · hardware

Assignment

Recorded

Dataset

AIPD

2023_r1 edition

Application

10418304

A computer-based feature recognition method is used to recognize features in a part of which are of a similar shape. A predetermined set of attributes are generated to represent a user-defined feature. This set is then used to produce an attributed graph which captures the feature shape. The graph is used in recognizing features of a shape similar to the feature shape in a solid model. To recognize a family of a shape feature with a variable number of faces, the shape feature is represented using a grammar employing symbols, rules and a terminating condition. Each symbol denotes an attributed face-adjacency graph and the rules denote permissible topological variations.

Machine learningVisionAI hardwareG06T 17/10G06V 10/426G06V 30/1985

AI classification

Vision1.00
Machine learning1.00
AI hardware0.64
Natural language0.14
Evolutionary computation0.13
Knowledge representation0.05
Planning0.02
Speech0.00

Ownership

GEOMETRIC SOFTWARE SOLUTIONS CO., LIMITED

assignment · 143530366

HCL TECHNOLOGIES LIMITED

assignment · 427680231

Assignors

VENKATARAMAN, SASHIKUMAR, SOHONI, MILIND A.

On an employer assignment, the assignors are typically the inventors.

© 2026 NYSGPT2525 LLC