Appearance-Invariant Signature for Graphical Object Classification

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Graphical object classification systems face reliability issues due to changes in appearance caused by transformations such as resolution changes, color inversion, and dynamic interactions, making template-based classification unreliable and inefficient.

Innovation Solution

The implementation of appearance-invariant signatures, generated from graphical object representations, allows for consistent classification across transformations by using attributes that remain stable despite changes in appearance, enabling classification systems to identify graphical objects accurately before and after transformations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If template-based classification is used, then classification can be performed using simple comparison methods, but classification reliability deteriorates when graphical objects undergo transformations such as resolution changes, color inversion, or dynamic interactions

Engineering Contradiction:

Improveclassification implementation simplicityVSAvoidclassification reliability under transformations

Core Design Contradiction:

Ease of manufactureVSReliability

Solution Approach 1:

The patent transforms graphical objects from spatial domain to frequency domain using Fourier transform, changing the representation parameters from pixel values to frequency components. This parameter transformation makes the classification features invariant to translation, scale, and rotation transformations, thereby maintaining classification reliability while keeping the method computationally efficient

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent replaces the traditional template matching mechanism with a frequency-domain correlation mechanism. Instead of directly comparing pixel values in spatial domain, the system uses Fourier transform to convert images to frequency domain, where correlation operations can efficiently handle transformations. This substitution maintains simplicity while improving robustness to transformations

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If extensive templates are generated to cover all possible transformations, then classification reliability improves, but system complexity and computational requirements increase significantly

Engineering Contradiction:

Improveclassification reliability under transformationsVSAvoidsystem complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent creates a universal classification method that handles multiple transformation types (translation, scale, rotation, color changes) through a single frequency-domain correlation framework. Instead of requiring separate templates for each transformation variant, the Fourier-based approach provides a unified solution that automatically adapts to different transformation types, significantly reducing system complexity

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

By transforming to frequency domain, the system extracts features that are inherently invariant to common transformations. The frequency representation naturally captures scale and rotation information, making the classification system robust without requiring extensive transformation-specific templates, thus reducing complexity while maintaining reliability

Inventive Principle:

Principle #35Parameter changes

3Adaptability or versatility

If traditional template matching is used, then the system can handle simple graphical objects, but it fails to maintain consistent classification when objects change appearance due to transformations

Engineering Contradiction:

Improvehandling of simple graphical objectsVSAvoidclassification consistency under appearance changes

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The patent applies Fourier transform to convert graphical objects from spatial domain to frequency domain, changing the representation parameters. This transformation enables the system to capture essential features that remain consistent under appearance changes such as translation, scale, and rotation, thereby maintaining classification consistency while still handling diverse graphical objects

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent replaces traditional spatial-domain template matching with frequency-domain correlation. This substitution allows the system to maintain adaptability to handle various graphical object types while achieving invariance to appearance changes through the mathematical properties of Fourier transform and correlation operations

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS10282059B2Graphical object appearance-invariant signature

Publication Date: 2019.05.07 MICRO FOCUS LLC

AI summary

In one implementation, a plurality of signature vectors from a multi-dimensional representation of a graphical object is generated. Each of the signature vectors comprises attributes that vary little in response to changes in shape, size, orientation, and visual layer appearance of the graphical object, and each of the signature vectors includes attributes based on operations of integration, differentiation, and transforms on the multi-dimensional representation of the graphical object. Each signature vector is composited into multiple portions from the plurality of signature vectors to define an appearance-invariant signature of the graphical object.