Face Detection Method Using Dual-Process Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Face detection methods in digital images face a trade-off between high detection probability and short processing time, making it challenging to achieve both simultaneously due to varying detection conditions such as inclination, orientation, size, and position of faces.
Innovation Solution
The method employs two detection processes: a first process with a high detection probability and longer processing time for favorable conditions, and a second process with a lower detection probability but shorter processing time for less favorable conditions, using two types of detectors and varying the position and size of the face to be detected based on predefined inclinations and positions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If detection conditions are specified finely to increase face detection probability, then detection probability improves, but processing time increases
Solution Approach 1:
The detection conditions are segmented into multiple inclination categories (first inclination with high detection probability expectation, and other inclinations). Different detection processes are applied to each segment, allowing fine specification for critical cases while using coarser specification for others, thus resolving the contradiction between detection probability and processing time.
Solution Approach 2:
The detection process dynamically adapts its granularity based on the face inclination. For faces with first inclination, a detailed detection process is applied; for other inclinations, a simplified process is used. This dynamic adjustment optimizes the balance between detection accuracy and processing speed across different scenarios.
2Productivity
If detection conditions are specified coarsely to reduce processing time, then processing time decreases, but detection probability decreases
Solution Approach 1:
Different detection condition specifications are applied locally based on face inclination characteristics. Rather than using a uniform coarse specification for all cases, the system applies fine specification locally to faces with first inclination (where high detection probability is expected) and coarser specification locally to other cases, thus maintaining high detection probability while reducing overall processing time.
Data Source
AI summary
When detecting a face included in an input image by use of at least two inclinations of the face to be detected, a first detection process is performed if the inclination of the face to be detected corresponds to a first inclination where a relatively high face detection probability is expected, and a second detection process having a relatively low face detection probability and a relatively short processing time compared with the first detection process is performed if the inclination of the face to be detected is other than the first inclination. When an input image is scanned to enable the change in face detection processes according to the position of the face to be detected, a different detection process is used according to the face detection probability of the position of the face to be detected.