Face Detection Image Selection Based on Expression Rarity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automatic imaging functions in cameras do not guarantee satisfactory images when facial expressions of multiple people are diverse, as the timing of shutter release is predetermined and does not account for varying expressions.
Innovation Solution
An image processing apparatus that detects faces, calculates evaluation values for facial expressions, and adjusts standards for extracting images based on the number of people and their expression values, allowing for automatic selection of images with the most satisfactory expressions, including options to prioritize expressions of children or those closest to the camera's center.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a predetermined shutter release timing is used in automatic imaging function, then the device operation is simple and fast, but the facial expression satisfaction in captured images deteriorates when multiple people have different expressions
Solution Approach 1:
The system performs preliminary detection of facial expressions and determination of the best capture timing before actually releasing the shutter. The face detection unit detects faces and the facial expression determination unit evaluates expressions in advance, allowing the system to select the optimal moment for capture rather than using fixed predetermined timing.
Solution Approach 2:
The system continuously monitors facial expressions of multiple people and uses this feedback to dynamically adjust the shutter release timing. The facial expression determination unit provides real-time feedback on expression states, enabling the control unit to determine the best timing based on current expression conditions rather than following a predetermined schedule.
2Device complexity
If facial expressions of multiple people are evaluated equally, then the processing is straightforward, but the image selection cannot prioritize expressions that a small number of people have
Solution Approach 1:
The system applies different evaluation weights to different facial expressions based on their rarity and importance. Rather than treating all expressions equally, the facial expression determination unit identifies expressions held by a small number of people and prioritizes these, allowing different parts of the expression evaluation process to have different qualities and priorities.
Solution Approach 2:
The system changes the evaluation parameters dynamically based on the detected facial expressions. When expressions varying among multiple people are detected, the system adjusts the selection criteria to prioritize images where a small number of people have distinctive expressions, rather than using a fixed equal-weight evaluation system.
3Ease of manufacture
If the shutter releases at fixed intervals, then the imaging function is simple to implement, but it does not guarantee capturing satisfactory facial expressions
Solution Approach 1:
The system transitions from static fixed-interval shutter release to dynamic timing based on real-time facial expression analysis. The control unit adjusts the shutter release timing dynamically according to the facial expressions detected by the face detection unit and evaluated by the facial expression determination unit, making the imaging function adaptive rather than rigid.
Solution Approach 2:
The system automatically determines the best shutter release timing based on its own facial expression detection and evaluation capabilities, without requiring external input or manual timing adjustment. The face detection unit and facial expression determination unit work together to self-regulate the imaging timing for optimal results.
Data Source
AI summary
An image processing apparatus includes a face detection unit which detects faces from an input image, an evaluation value calculation unit which calculates an evaluation value expressing a degree of a facial expression for each of the facial expressions of the faces detected by the face detection unit, and a control unit which changes a standard for extracting an image such that an image including the facial expressions, that a small number of people have, is easily extracted, based on the number of people for each of the facial expressions of the faces detected by the face detection unit and an evaluation value for each facial expression of the face calculated by the evaluation value calculation unit.