Out-of-Focus Blurring Evaluation Using Edge Pixel Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for evaluating out-of-focus blurring in images, especially those with blurred edges, are inaccurate due to the difficulty in visually assessing small liquid crystal display screens and the challenge of distinguishing between in-focus and out-of-focus areas.
Innovation Solution
The method involves extracting edge pixels from image information, calculating the number of pixels intersecting with the boundary, and performing out-of-focus blurring evaluation based on these calculations, including applying a Sobel filter to suppress noise and determine edge widths, which allows for accurate evaluation even in images with blurred contours.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If visual evaluation is performed on a liquid crystal display section, then image quality assessment is possible, but the small screen size makes it difficult to accurately evaluate out-of-focus blurring
Solution Approach 1:
The patent replaces visual evaluation (human eye observation on display) with automatic computer-based image processing. The system uses edge detection algorithms and luminance difference calculations to objectively measure out-of-focus blurring, eliminating the limitations of small display screens and human visual assessment.
Solution Approach 2:
The patent introduces intermediate calculation steps including edge pixel extraction, boundary detection, and luminance difference computation. These intermediary processes transform the raw image data into measurable metrics that accurately represent out-of-focus conditions without requiring direct visual inspection.
2Measurement precision
If edge detection is performed on images with blurred contours, then out-of-focus evaluation can be attempted, but the gentle luminance change around edges makes accurate detection difficult
Solution Approach 1:
The patent applies different processing strategies to different regions of the image. It identifies edge pixels through gradient analysis and applies specific luminance difference calculations only to boundary regions, while handling interior pixels differently. This localized approach improves edge detection accuracy without being affected by gentle luminance changes in non-edge areas.
Solution Approach 2:
The patent performs preliminary edge detection and boundary identification before conducting the main out-of-focus evaluation. By pre-identifying edge pixels and their boundaries using gradient-based methods, the system prepares accurate reference data that facilitates subsequent blurring measurement even in challenging images with indistinct contours.
Data Source
AI summary
The present method of evaluating image information, includes: extracting, from image information to be evaluated, a plurality of edge pixels located in a boundary of an image expressed by the image information; calculating, for each of the edge pixels, a number of pixels that include the edge pixel targeted for calculation, that exist in the boundary, and that are arranged in a direction intersecting with the boundary; and performing out-of-focus blurring evaluation of the image information on the basis of the number of pixels that exist in the boundary and a number of the edge pixels.