Anisotropic Diffusion Illumination Normalization

Solution Overview

Problem

Conventional illumination normalization techniques struggle to effectively remove spurious edges caused by illumination variations in digital images without affecting real edges, leading to reduced aesthetic quality and difficulties in image processing algorithms like facial recognition.

Innovation Solution

The method employs anisotropic diffusion processing, including model-based anisotropic diffusion, to differentiate between real and spurious edges, using a processor and memory to perform image processing and generate a reflectance estimate by isolating and removing spurious edges while preserving real edges.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Object-affected harmful factors

If conventional illumination normalization techniques (level compression, histogram stretching) are used, then illumination variations are reduced, but spurious edges cannot be effectively removed and global intensity variations are introduced

Engineering Contradiction:

Improveillumination variationsVSAvoidedge preservation accuracy

Core Design Contradiction:

Object-affected harmful factorsVSManufacturing precision

Solution Approach 1:

The patent applies local quality by using anisotropic diffusion with directionally-dependent conductivity that adapts to local image features. The diffusion process selectively smooths illumination variations in certain directions while preserving edges in orthogonal directions, allowing different parts of the image to be treated with different diffusion strengths based on local edge orientation and intensity gradients.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent changes parameters by dynamically adjusting diffusion coefficients based on local image characteristics such as gradient magnitude and orientation. The conductivity tensor parameters are modified according to local edge detection results, enabling the system to adaptively control diffusion strength and direction to remove spurious edges while preserving real edges.

Inventive Principle:

Principle #35Parameter changes

2Object-affected harmful factors

If strong diffusion processing is applied to remove spurious edges, then illumination normalization improves, but real edges may be blurred or lost

Engineering Contradiction:

Improvespurious edgesVSAvoidreal edge information

Core Design Contradiction:

Object-affected harmful factorsVSLoss of information

Solution Approach 1:

The patent applies local quality by using anisotropic diffusion with directionally-dependent conductivity that adapts to local image features. The diffusion process selectively smooths illumination variations in certain directions while preserving edges in orthogonal directions, allowing different parts of the image to be treated with different diffusion strengths based on local edge orientation and intensity gradients.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent implements feedback by using edge detection results and gradient analysis to continuously adjust diffusion coefficients during the processing. The system monitors local image characteristics and modifies diffusion strength accordingly, reducing diffusion near detected edges to preserve real edge information while maintaining diffusion in regions with spurious edges.

Inventive Principle:

Principle #23Feedback

3Ease of manufacture

If conventional techniques are used, then processing simplicity is maintained, but ability to handle sharp spurious edges is insufficient

Engineering Contradiction:

Improveprocessing simplicityVSAvoidspurious edge removal accuracy

Core Design Contradiction:

Ease of manufactureVSMeasurement precision

Solution Approach 1:

The patent changes parameters by dynamically adjusting diffusion coefficients based on local image characteristics such as gradient magnitude and orientation. The conductivity tensor parameters are modified according to local edge detection results, enabling the system to adaptively control diffusion strength and direction to remove spurious edges while preserving real edges.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent applies dynamics by making the diffusion process adaptive rather than static. The diffusion coefficients and conductivity tensor are dynamically adjusted based on local image analysis, allowing the system to respond to varying edge characteristics and illumination patterns throughout the image, thereby improving spurious edge removal accuracy.

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS7747045B2Method and apparatus for diffusion based illumination normalization

Publication Date: 2010.06.29 FUJIFILM CORP

AI summary

An exemplary illumination normalization method is provided which includes receiving an input image having at least one spurious edge directly resulting from illumination, performing anisotropic diffusion on the input image to form a diffusion image, and removing at the least one spurious edge using the diffusion image. Another embodiment consistent with the invention is an apparatus for performing illumination normalization in an image which includes a processor operably coupled to a memory storing input image data which contains an object of interest having at least one spurious edge directly resulting from illumination, a model of a representative object of interest, and functional processing units for controlling image processing, wherein the functional processing units further include a model based anisotropic diffusion module which predicts edge information regarding the object of interest based upon the model, and produces a reflectance estimation utilizing the predicted edge information.