Red Eye Detection Using Edge and Red Pixel Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional redeye image detection systems fail to accurately detect and compensate redeye images when only one eye or no complete human face is present in the image, and they are not effective in detecting animal eyes, due to complex computations and limited functionality.
Innovation Solution
A redeye image detection method and apparatus that simultaneously uses red pixel and edge detection techniques to identify redeye images, including edge detection circuits, red pixel detection circuits, and determining circuits to derive preliminary detection results and compensate for redeye images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional facial detection and complexion determination processes are used to detect redeye images, then redeye detection can be performed, but the system complexity increases and detection fails when complete human face is not present
Solution Approach 1:
The patent extracts the essential redeye detection function from the complex facial detection and complexion determination processes. By directly detecting red pixels and edge characteristics in the image, the system eliminates the need for complete facial detection and multiple computation processes, thereby reducing system complexity while maintaining detection reliability even when complete human faces are not present
Solution Approach 2:
The patent segments the redeye detection process into independent red pixel detection and edge detection operations. This segmentation allows the system to detect redeye images without requiring complete facial detection, as each operation can function independently to identify red eye characteristics even when only partial facial features are present
2Adaptability or versatility
If conventional redeye detection is limited to human eyes, then detection process is simple, but animal eyes cannot be detected
Solution Approach 1:
The patent makes the detection system universal by using red pixel detection and edge detection that can identify eyes of any species. The detection process does not distinguish between human and animal eyes, allowing the same simple algorithm to detect redeye images from both humans and animals, thereby expanding detection scope without increasing complexity
3Measurement precision
If multiple tedious computations are performed to derive facial image, then facial image can be derived correctly, but computation time increases
Solution Approach 1:
The patent extracts only the essential red pixel and edge detection operations from the complex facial image derivation process. By removing unnecessary complexion determination and other tedious computations, the system achieves sufficient accuracy for redeye detection with significantly reduced computation time, eliminating the trade-off between accuracy and time
Data Source
AI summary
A redeye image detection method for detecting at least one redeye image in an input image includes: deriving an edge detection result by performing an edge detection process to detect an edge characteristic of the input image; deriving a red pixel detection result by performing a red pixel detection process to detect red pixels in the input image, and utilizing a determining circuit to derive a preliminary detection result according to the edge detection result and the red pixel detection result, and derive the redeye image according to the preliminary detection result.