Markov Chain Blur Measurement via Transition Probabilities
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for measuring image blur in digital images and sequences are computationally complex and often rely on assumptions or reference information, leading to inaccurate estimates, especially when sharp edges are absent or when dealing with compressed images.
Innovation Solution
A computationally efficient method using a Markov Chain to calculate a gradient image array and construct a transition probability matrix from pixel data, allowing for quantitative image blur measurement without reference information, by pooling transition probability data to determine the degree of blur.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If spatial domain edge detection is used to measure image blur, then the measurement can be performed directly on image data, but the accuracy deteriorates when sharp edges are absent
Solution Approach 1:
The patent introduces transition probability matrices as an intermediary mechanism that bridges the gap between image data and blur measurement. Instead of directly analyzing edges (which may be absent), the system computes transition probabilities between adjacent pixel intensity values, which serve as a mediator to infer blur characteristics even when traditional edges are not present in the image.
Solution Approach 2:
The patent transforms the measurement approach by changing from direct spatial domain edge analysis to a probability-based parameter system. By computing transition probabilities from the image data and using these probabilities to estimate blur, the system changes the fundamental parameter being measured from edge sharpness to probability distribution characteristics, enabling accurate blur measurement even without sharp edges.
2Measurement precision
If power spectrum analysis is used to measure image blur, then the measurement can capture frequency information, but the computational complexity increases significantly
Solution Approach 1:
The patent extracts only the essential information needed for blur measurement from the image data, specifically the transition probabilities between adjacent pixel values. By taking out only this critical probability information and discarding redundant frequency domain computations, the system achieves accurate blur measurement while significantly reducing computational complexity compared to full power spectrum analysis.
Solution Approach 2:
Instead of performing complete power spectrum analysis which examines all frequency components, the patent applies partial action by focusing only on the transition probability characteristics of adjacent pixel values. This partial approach extracts sufficient blur information without the excessive computational burden of analyzing the entire frequency spectrum.
3Measurement precision
If DCT coefficient histogram analysis is used to measure image blur, then the measurement can account for compression effects, but the computational complexity remains high
Solution Approach 1:
The patent extracts blur measurement capability from the complex DCT coefficient histogram analysis. Instead of analyzing the entire DCT coefficient distribution, the system extracts only the transition probability information from adjacent pixel values in the spatial domain, which suffices to capture compression effects and provide accurate blur measurement with reduced computational complexity.
Data Source
AI summary
Systems and methods of performing quantitative measurements of image blur in digital images and digital image sequences that are computationally efficient, and that employ no reference information in the measurement process. Each of the image blur measurements is performed using a Markov Chain, where a gradient image array is calculated for a pixel array derived from a given digital image, and a transition probability matrix is constructed for the transition probabilities between adjacent elements in the gradient image array. The transition probability data contained in the transition probability matrix can be pooled or otherwise accumulated to obtain a quantitative measurement of image blur in the given digital image.