Red Eye Detection Using Person-Specific Parameters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing red eye detection methods in digital image processing suffer from reduced accuracy due to uniform detection properties applied across all facial images, leading to missed detections in individuals prone to red eye occurrence and false positives in those not prone to it.
Innovation Solution
A method, apparatus, and program that detect facial images and discriminate between individuals registered in a database, using red eye detecting parameters tailored to each person, allowing for customized detection and correction settings based on individual differences in red eye frequency and color.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If uniform red eye detection parameters are applied to all facial images, then the detection process is simple and fast, but detection accuracy deteriorates due to individual differences in red eye occurrence frequency and color characteristics
Solution Approach 1:
The patent applies local quality by using different red eye detection parameters for different individuals. Instead of uniform parameters for all faces, the system stores and applies person-specific detection parameters (such as color thresholds and detection sensitivity) that are tailored to each individual's characteristics, thereby improving detection accuracy while managing complexity through selective personalization.
Solution Approach 2:
The patent implements parameter changes by adjusting red eye detection parameters based on individual characteristics. The system changes detection parameters (such as color range thresholds, sensitivity levels, and detection criteria) according to stored profiles of different persons, allowing optimal detection for each individual while maintaining a systematic approach through pre-stored parameter sets.
2Reliability
If uniform detection parameters are used for all individuals, then the system is easy to operate, but false positive detections occur in individuals not prone to red eye while missed detections occur in those prone to red eye
Solution Approach 1:
The patent applies preliminary action by pre-storing detection parameters for multiple individuals in a database before actual red eye detection occurs. The system prepares person-specific detection profiles in advance, including color characteristics and detection thresholds, so that when detection is needed, the appropriate parameters are already available, improving reliability without adding operational complexity during the detection process.
Data Source
AI summary
During red eye processing, whether a facial image detected from within an entire image is that of a specific person registered in a database is judged. In the case that it is judged that the facial image is of the specific person registered in the database, red eye detection is performed within the facial image, employing red eye detecting parameters, which are also registered in the database.