Mobile Camera Eye Blink Detection via Down-Sampled Preview Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional digital cameras are unable to efficiently capture images by filtering out eye blink patterns and introducing face smile patterns, resulting in unsatisfactory image presentation.
Innovation Solution
An image capture method and mobile camera device that utilize an image sensor and image signal processor to detect eye blink and face smile patterns in preview images of lower resolution, allowing for efficient filtering of eye blink patterns and optional introduction of face smile patterns in the result image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If eye blink detection is performed on full-resolution original images, then detection accuracy is improved, but processing time and computational load increase significantly
Solution Approach 1:
The patent divides the image processing task into two stages: first processing a down-sampled preview image to detect eye blink patterns, then selectively processing only the original high-resolution image if eye blinks are detected. This segmentation allows most images to be processed quickly while maintaining accuracy when needed.
Solution Approach 2:
The patent introduces a preview image as an intermediary between the captured original image and the final processing step. This preview image serves as a low-cost proxy for detecting eye blink patterns, allowing the system to avoid expensive full-resolution processing for images without eye blinks.
2Manufacturing precision
If multiple consecutive images are captured to filter eye blinks, then image quality is improved, but capture time and buffer memory requirements increase
Solution Approach 1:
The patent performs preliminary detection of eye blink patterns on down-sampled preview images before committing to capture and process multiple full-resolution images. This preliminary action allows the system to determine in advance whether multiple captures are necessary, reducing unnecessary capture time.
Solution Approach 2:
The patent captures a consecutive plurality of second original images only when eye blink patterns are detected in the preview image, rather than always capturing multiple images. This partial action approach applies the excessive capture strategy only when needed, balancing image quality with capture time efficiency.
3Reliability
If eye blink detection is performed on all captured images, then reliability of filtered images is improved, but processing complexity and computational resources increase
Solution Approach 1:
The patent introduces preview images as an intermediary filtering stage that reduces the number of full-resolution images requiring detailed eye blink detection. This two-stage approach maintains reliability by ensuring that only images passing the preview filter undergo full detection, reducing overall processing complexity.
Solution Approach 2:
The patent extracts and processes only the necessary information (eye blink patterns) from down-sampled preview images, then uses this extracted information to guide further processing of original images. This extraction approach avoids redundant processing while maintaining detection reliability.
Data Source
AI summary
By grading preview images of original images for the purpose of filtering eye blink patterns off and/or introducing face smile patterns, a result image can be generated with least eye blink patterns and most face smile patterns for satisfying image quality.