Nonlinear Blur Estimation via Pixel Intensity Distribution Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image processing technologies cannot effectively estimate or eliminate image blur caused by defocus due to the use of Gaussian point spread functions, which are inadequate for nonlinear imaging systems, resulting in suboptimal image quality.
Innovation Solution
A method that involves transmitting an input image to an image processing device to produce a synthesized blur image using a nonlinear image sensing function, matching pixel intensity distribution, and calculating a blur degree parameter to provide accurate blur estimation and evaluation of image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If Gaussian point spread function is used for blur estimation, then the processing is simple and fast, but the estimation accuracy is insufficient for nonlinear imaging systems
Solution Approach 1:
The patent changes the fundamental parameter of the point spread function from Gaussian to a non-uniform distribution function that models actual defocus blur characteristics. This parameter change enables accurate representation of blur in nonlinear imaging systems while maintaining computational feasibility through efficient algorithm design.
Solution Approach 2:
The patent introduces dynamic adaptation by using image gradient information to guide the blur estimation process. The estimation algorithm dynamically adjusts to local image characteristics, allowing accurate blur measurement in nonlinear systems without requiring complex fixed-model approaches.
2Ease of manufacture
If current image modification technologies are used, then the processing method is simple, but the image blur cannot be eliminated effectively
Solution Approach 1:
The patent replaces traditional mechanical-style Gaussian filtering with a physics-based model that substitutes the inappropriate Gaussian assumption with a more accurate non-uniform distribution model. This substitution maintains processing simplicity while dramatically improving blur elimination effectiveness by matching the actual optical physics of defocus.
Solution Approach 2:
The patent changes the core parameter model from Gaussian distribution to non-uniform distribution, enabling effective blur elimination in nonlinear imaging systems while keeping the processing framework relatively simple through algorithmic efficiency.
3Adaptability or versatility
If Gaussian PSF is adopted for modifying blur, then the linear system model is maintained, but the nonlinear sensing device characteristics cannot be represented
Solution Approach 1:
The patent fundamentally changes the PSF parameter from Gaussian to a non-uniform distribution that captures nonlinear sensing device characteristics. This parameter change enables the model to adapt to actual device behavior while maintaining computational tractability through efficient implementation.
Solution Approach 2:
The patent segments the imaging model into distinct components: the non-uniform PSF for defocus modeling and separate handling of nonlinear sensing characteristics. This segmentation allows each component to be optimized independently, achieving high accuracy for nonlinear devices without overwhelming complexity.
Data Source
AI summary
A method for estimating blur degree of image and a method for evaluating image quality are revealed. First, an input image is transmitted to an image processing device for producing a synthesized blur image including a nonlinear image sensing function according to a pixel intensity distribution parameter of the input image. Next, the image processing device matches the pixel intensity distribution according to the input image and the synthesized blur image for producing a blur degree parameter; By the way, the image processing device further estimating an estimated blur result according to the blur degree parameter. The method for estimating blur degree of image can be further applied to estimate blur distribution for a plurality of regions of interest of a plurality of input images. Thereby, the blur distribution of the input images can be estimated, and thus further evaluating the image quality of the plurality of input images.