Video Quality Estimation Without Reference Images

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional video quality metrics, such as PSNR and SSIM, are inadequate for scenarios lacking uncompressed references, requiring stabilization, contrast enhancement, sharpening, super-resolution, or dealing with atmospheric conditions like fog and haze, and fail to account for camera settings and motion.

Innovation Solution

A video interpretability and quality estimation system that generates estimates using feature characteristics, including ground sample distance, relative edge response, peak signal to noise ratio, camera motion, contrast, and artifacts, even in the absence of reference images, employing equations and techniques like phase congruency and perceptual analysis windows to assess video quality across various conditions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional video quality metrics (PSNR, SSIM) are used, then measurement simplicity is maintained, but measurement precision deteriorates in scenarios lacking uncompressed references or involving atmospheric conditions

Engineering Contradiction:

Improvevideo quality assessment accuracyVSAvoidsystem complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the video quality assessment into multiple independent feature extraction components (spatial frequency analysis, edge response measurement, contrast evaluation, artifact detection) that can be processed separately and combined. This allows comprehensive quality assessment without requiring complex reference-based comparisons, thereby improving measurement precision while managing system complexity through modular design

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces an intermediary computational model that estimates video quality metrics by analyzing intermediate features (spatial frequencies, edge responses, contrast values) rather than directly comparing with reference images. This intermediary approach enables accurate quality assessment in scenarios where conventional direct comparison methods fail, improving precision without requiring uncompressed references

Inventive Principle:

Principle #24Intermediary (Mediator)

2Adaptability or versatility

If reference images are required for quality assessment, then measurement precision improves, but adaptability deteriorates in scenarios where references are unavailable

Engineering Contradiction:

Improveassessment scenario flexibilityVSAvoidquality estimation accuracy

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent implements a self-service quality assessment mechanism that extracts all necessary quality indicators directly from the video content itself without requiring external reference images. The system performs spatial frequency analysis, edge response measurement, and artifact detection autonomously on the input video, enabling adaptability to scenarios where references are unavailable while maintaining measurement precision through multiple content-based features

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent creates a universal quality assessment system that can operate in multiple scenarios (with references, without references, with atmospheric conditions, without atmospheric conditions) by implementing a multi-functional feature extraction framework. The same core algorithms serve different assessment needs, providing both adaptability across scenarios and consistent measurement precision through unified processing

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Measurement precision

If comprehensive feature analysis is performed, then measurement precision improves, but computation time increases

Engineering Contradiction:

Improvequality metric accuracyVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent performs preliminary action by pre-defining and pre-configuring the feature extraction algorithms (spatial frequency filters, edge detection operators, contrast calculation methods) before actual video processing. This preparation allows the system to execute comprehensive quality analysis with optimized computational paths during runtime, improving measurement precision while reducing actual processing time through advance setup

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent implements partial action by selectively applying different levels of feature analysis based on assessment needs. The system can perform complete comprehensive analysis when high precision is required, or apply simplified versions of the algorithms for faster processing when time is constrained, allowing flexible trade-off between measurement precision and computation time

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS9858656B2Video interpretability and quality estimation

Publication Date: 2018.01.02 RAYTHEON CO

AI summary

According to embodiment, an interpretability and quality estimation method includes generating an interpretability or quality estimate for an image or frame using one or more feature characteristics of a reference image when a reference image associated with an image is available or partially available, and generating an interpretability or quality estimate for the image or frame using one or more feature characteristics of the image when a reference image associated with the image is not available.