Face Detection Area Correction for Sideways Orientation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing systems for autofocus in camcorders face challenges in accurately detecting and focusing on human faces, especially when the face is oriented sideways, leading to misalignment of focus detection areas and potential focusing on the background rather than the face.
Innovation Solution
An image processing apparatus that includes a first detection unit for detecting a face area, a second detection unit for estimating the face area based on human body detection, a correction unit to adjust the face detection area, and a control unit for focus detection, exposure, and color control, ensuring accurate focus on the face by reducing background influence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If face detection is performed based on eye position, then face area can be detected, but detection accuracy decreases when the face is oriented sideways
Solution Approach 1:
The patent combines face detection results from multiple detection methods (eye position-based detection and alternative detection methods) to determine the final face area. When the face is oriented sideways, the system merges information from different detection approaches to accurately identify the face area, thereby maintaining detection accuracy across various face orientations.
2Reliability
If focus detection area includes the detected face area, then autofocus can be performed on the face, but the background may be included in the focus detection area causing the background to be in focus instead
Solution Approach 1:
The patent applies different processing approaches to different regions within the face area. By identifying and excluding background regions from the focus detection area while maintaining the face area, the system ensures that focus detection is performed locally on the face subject rather than being influenced by background elements, thereby improving focus detection accuracy.
3Device complexity
If a single face detection method is used, then the system is simple, but detection accuracy decreases when the face orientation varies
Solution Approach 1:
The patent implements a multi-functional face detection system that can handle various face orientations using a unified detection framework. The system incorporates multiple detection methods that can be adaptively applied based on the detected face orientation, making the detection system universal across different scenarios while maintaining accuracy.
Data Source
AI summary
At least either of a location and a size of a subject area that is detected from an image by a first detection unit is corrected based on information about an area corresponding to the subject area detected by the first detection unit, the area being included in an area that is detected from the image by a second detection unit and in which the subject area can be estimated. A particular subject area can be detected in an image with higher accuracy.