Automatic Red Eye Removal Using Segmented Bayesian Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for red eye removal in photographs require user intervention or have high computational complexity, and existing automated methods are inefficient in processing and memory usage.
Innovation Solution
A method that identifies potential red eye pixels in low resolution images, groups them into regions, determines working areas, applies Bayesian classifiers to high resolution images, and replaces red eye pixels with black pixels, reducing computational steps and memory usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If automated red eye removal is implemented, then user intervention is eliminated, but computational complexity increases
Solution Approach 1:
The patent segments the red eye removal process into distinct modules: detection module that identifies red eye regions, classification module that categorizes detected regions, and removal module that applies correction. This segmentation allows each module to handle specific tasks efficiently, reducing overall computational complexity while maintaining full automation.
Solution Approach 2:
The patent applies preliminary filtering and classification before the actual red eye removal operation. By pre-identifying and categorizing potential red eye regions using color space analysis and shape criteria, the system prepares data structures and selects appropriate removal algorithms in advance, reducing the computational burden during the execution phase.
2Measurement precision
If full resolution image processing is applied, then red eye removal accuracy is improved, but memory usage increases
Solution Approach 1:
The patent processes the image by segmenting it into color channels (RGB or HSV spaces) and further segments potential red eye regions for focused analysis. This segmentation allows the system to apply intensive processing only to relevant portions of the image at full resolution, while other areas are handled more efficiently, optimizing the balance between accuracy and memory consumption.
3Manufacturing precision
If multiple processing steps are applied, then red eye removal quality is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary detection and classification of red eye regions using color constraints and shape criteria before applying the actual removal algorithm. This preliminary action prepares the data structures and selects the appropriate removal method in advance, allowing the main processing step to execute more efficiently with reduced iteration and reprocessing.
Solution Approach 2:
The patent applies partial processing by focusing computational resources only on identified red eye regions rather than processing the entire image uniformly. The classification module determines the appropriate level of processing for each detected region, applying more intensive algorithms only where necessary to achieve quality removal while minimizing overall processing time.
Data Source
AI summary
A method for removing red eyes includes (1) identifying potential red eye pixels in a low resolution image (hereafter “LR red eye pixels”), (2) grouping contiguous LR red eye pixels into red eye regions (herafter “LR red eye regions”), (3) determining working areas around the LR red eye regions (hereafter “LR working areas”), (4) determining classifiers from the LR working areas, (5) retrieving working areas from a high resolution image that correspond to the LR working areas (hereafter “HR working areas”), (6) applying the classifiers to the HR working areas to identify potential red eye pixels (hereafter “HR red eye pixels”), (7) grouping contiguous HR red eye pixels into red eye regions (hereafter “HR red eye regions”), (8) determining some of the HR red eye regions as the red eyes, and (9) replacing the HR red eye pixels in the red eyes with black pixels.