Visual Quality Metrics via Spatial Shifting for Blindness Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current image quality assessment methods face challenges in achieving high accuracy while maintaining low computational complexity, particularly in detecting distortions that do not affect the average pixel intensity within blocks, leading to 'blindness' in certain types of image degradation.

Innovation Solution

The method employs spatial shifting of images by predefined amounts to make patterns undetectable during transformation, using techniques like Haar or Daubechies transforms, and computes quality metrics by comparing shifted and unshifted versions of distorted and reference images, thereby increasing the robustness of visual quality metrics.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If spatial shifting is applied to detect certain distortion patterns, then measurement precision improves, but device complexity increases

Engineering Contradiction:

Improvequality assessment accuracyVSAvoidcomputational complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The image is divided into multiple blocks, and the transformation is applied to each block separately. This segmentation allows the system to detect local distortion patterns that would be missed in a global transformation, improving measurement precision without requiring a completely complex new system architecture

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The method applies spatial shifting and multiple transformations in advance before final quality assessment. By performing these preprocessing steps beforehand, the system prepares the data to reveal hidden distortion patterns, improving detection accuracy while organizing the computational workload in a manageable sequence

Inventive Principle:

Principle #10Preliminary action

2Reliability

If multiple transformations are applied to increase robustness, then reliability improves, but loss of time increases

Engineering Contradiction:

Improverobustness of quality metricVSAvoidcomputational time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The computational process is divided into distinct transformation stages applied to image blocks. This segmentation allows parallel processing of different blocks and transformations, reducing overall computational time while maintaining the reliability benefits of multiple transformations

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The method applies transformations at multiple levels and with different parameters, performing more transformations than a single basic transform would provide. This partial/excessive action ensures robustness against various distortion types while the systematic organization prevents excessive time consumption through efficient resource utilization

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS8983206B2Method and system for increasing robustness of visual quality metrics using spatial shifting

Publication Date: 2015.03.17 ECOLE DE TECH SUPERIEURE

AI summary

A visual quality assessment method and system are based on deriving a quality metric by comparing sub-band approximations of a distorted image and an undistorted version of the same image, providing a good compromise between computational complexity and accuracy. The sub-band approximations are derived from Discrete Wavelet (Haar) transforms of small image blocks of each image. Due to inherent symmetries, the wavelet transform is “blind” to certain types of distortions. But the accuracy of the method is enhanced, and the blindness of the transform is overcome, by computing quality metrics for the distorted image as well as computing quality metrics for a shifted version of the distorted image and combining the results.