Face Detection via Horizontal Edge Averaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing face detection technologies face high computational loads and require numerous pre-prepared face models due to variations in face size and orientation, making them inefficient for real-time face direction detection.
Innovation Solution
A face detection method that calculates the central location of a face by detecting horizontal edges and determining their average value, using edge abstraction and time-differentiated images to reduce computational load and eliminate noise, without the need for extensive face model data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If template matching methods are used to detect face direction, then face direction detection can be achieved, but the computational load becomes enormous due to the need to prepare and match multiple templates of various sizes and orientations
Solution Approach 1:
The invention extracts only the essential horizontal edge information from the complex task of complete face template matching. By focusing solely on horizontal edges with predetermined vertical lengths and calculating their average horizontal location, the method extracts the minimal necessary data to determine face direction without requiring comprehensive template libraries, thus dramatically reducing computational load while maintaining detection accuracy.
Solution Approach 2:
The invention segments the face detection task into a specific subset - detecting only horizontal edges with predetermined vertical lengths. This segmentation allows the system to ignore irrelevant edge information and focus solely on the critical features needed for face direction determination, reducing the search space and computational requirements significantly.
2Device complexity
If flat templates are used for face orientation estimation, then the process can be simplified, but the accuracy deteriorates because human faces with diverse hair styles and facial lines do not match flat templates well
Solution Approach 1:
The invention extracts the invariant horizontal edge features that remain consistent across different face types, hair styles, and orientations. By focusing on horizontal edges with predetermined vertical lengths, the method captures the essential geometric information needed for orientation detection without being affected by variable facial features, achieving both simplicity and accuracy.
Solution Approach 2:
The invention applies local quality by selecting edges with specific predetermined vertical lengths that correspond to characteristic facial features. This localized approach ensures that only relevant edge information is used for detection, maintaining high accuracy while avoiding the complexity of comprehensive template matching.
3Measurement precision
If multiple face models are prepared to account for individual differences, then detection accuracy can be maintained, but the amount of processing increases significantly
Solution Approach 1:
The invention extracts the universal horizontal edge detection mechanism that works across all individuals without requiring person-specific models. By detecting horizontal edges with predetermined vertical lengths and calculating their average horizontal location, the method achieves accurate face direction detection for all individuals using a single, universal approach, eliminating the need for multiple pre-prepared face models and significantly reducing processing time.
Data Source
AI summary
A central location of a face detecting device comprises an image input means for inputting a captured face image, a horizontal edge detecting means for detecting a horizontal edge having a predetermined vertical length based on the face image inputted by the image input means; and a central location of a face detecting means for calculating an average value of a horizontal location data of the horizontal edge detected by the horizontal edge detecting means, and determining the average value as a central location of the face.