Image Sharpness Detection Using Frequency-Spectrum Envelope Fitting

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image sharpness determination methods are dependent on image content and require significant training data, making them inefficient for real-time decisions in dynamically changing scenes.

Innovation Solution

A method using a one-dimensional frequency spectrum analysis, fitting a straight line to characterize image sharpness through the envelope's smoothness, allowing scene-independent and computationally efficient real-time sharpness assessment.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If machine learning tools are used to determine image sharpness, then classification accuracy can be improved, but training data requirements and processing time increase significantly

Engineering Contradiction:
Improvesharpness determination accuracyVSAvoidtraining data processing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent extracts the essential characteristic of sharpness (envelope smoothness in frequency domain) from complex image content, separating the measurement task from content-dependent processing. This extraction enables a simplified, content-independent approach that achieves accurate sharpness determination without requiring extensive training data or complex machine learning models.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent replaces the mechanical system of machine learning training and classification with a mathematical transformation approach (frequency domain analysis). By substituting iterative learning processes with direct spectral analysis and envelope fitting, the system achieves comparable or superior accuracy with significantly reduced computational time and data requirements.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If edge detection methods are used to determine sharpness, then measurement precision can be improved, but dependency on image content increases

Engineering Contradiction:
Improvesharpness measurement accuracyVSAvoidscene independence
Core Design Contradiction:
Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent transitions from spatial domain edge detection to frequency domain analysis. By transforming the image to the frequency domain and analyzing the envelope smoothness of spectral components, the method measures sharpness characteristics without being influenced by scene content, objects, or spatial arrangements. This dimensional transformation enables content-independent sharpness assessment.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Solution Approach 2:

The patent changes the measurement parameter from edge-based spatial features to frequency domain envelope characteristics. By measuring the smoothness of the spectral envelope rather than counting edges or detecting spatial transitions, the system achieves sharpness determination that is invariant to image content, enabling consistent performance across diverse scenes.

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If complex sharpness analysis methods are used, then measurement precision can be improved, but computational complexity increases

Engineering Contradiction:
Improvesharpness determination accuracyVSAvoidcomputational resource requirements
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the sharpness measurement task into distinct frequency domain operations: transformation to frequency domain, envelope extraction, and smoothness evaluation. This segmentation allows each operation to be performed efficiently with well-established algorithms (FFT, Hilbert transform), reducing overall computational complexity while maintaining measurement precision.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the measurement approach from content-dependent spatial analysis to frequency domain envelope analysis. This parameter change enables the use of efficient spectral estimation techniques that require fewer computational resources compared to complex machine learning models or exhaustive edge detection algorithms, while achieving comparable or superior accuracy.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4338125B1Method, data processing system, computer program product and computer readable medium for determining image sharpness
Publication Date: 2025.07.02 AIMOTIVE KFT
  • EP4338125B1 patent drawingFigure 1~3B
  • EP4338125B1 patent drawingFigure 4~5
  • EP4338125B1 patent drawingFigure 6~7

AI summary

The invention is a method for determining whether an input image is sufficiently sharp, comprising the steps of - providing a blur score threshold value, - inputting the input image to an image processing system, - generating, by the image processing system, a two-dimensional frequency spectrum of the input image, - generating a one-dimensional frequency spectrum (14) from the two-dimensional frequency spectrum, - fitting a straight line (16) on the one-dimensional frequency spectrum (14), and - determining a blur score value based on a residual of the fitting, and considering the input image as sufficiently sharp based on a comparison of the blur score value with the blur score threshold value. The invention further relates to a data processing system, a computer program product and a computer readable medium carrying out the above method.