Dual Image Sensor Gaze Detection for Region of Interest
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image processing systems in smartphones lack the ability to automatically adjust settings such as auto white balance, auto exposure, and auto focus for a rear camera based on input from a front camera, and cannot efficiently detect the region of interest desired by the user.
Innovation Solution
An image processing system that includes a first image sensor for the rear camera and a second image sensor for the front camera, where the second image sensor generates probability maps representing the user's gaze direction and face angles, allowing the image processor to detect the region of interest and adjust the settings accordingly, enabling the rear camera to perform functions like auto white balance, auto exposure, and auto focus on that region.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple image sensors are used in a smartphone (front camera and rear camera), then the functionality and versatility of the imaging system is improved, but the device complexity and power consumption increase
Solution Approach 1:
The imaging system is segmented into two distinct image sensors: a front camera (second image sensor) for capturing user face and gaze information, and a rear camera (first image sensor) for capturing the region of interest. This segmentation allows each sensor to specialize in specific functions, improving overall system versatility while managing complexity through functional division
Solution Approach 2:
The front camera serves multiple functions: it captures user face images for gaze detection, generates probability maps for eye direction, and provides information for automatic region of interest detection in the rear camera. This multi-functionality improves system adaptability without requiring additional dedicated sensors for each function
2Measurement precision
If the front camera operates in normal mode continuously, then the accuracy of gaze detection is improved, but the power consumption increases
Solution Approach 1:
The front camera operates dynamically in two modes: low power mode for normal operation and normal mode when gaze detection is required. The system transitions between these modes based on operational needs, maintaining measurement precision when necessary while minimizing power consumption during routine operations
Solution Approach 2:
The front camera captures images periodically in low power mode and switches to normal mode only when gaze direction detection is required. This periodic activation of high-precision mode reduces overall power consumption while maintaining the capability for accurate gaze detection when needed
3Ease of operation
If manual adjustment of image capture settings is required, then the ease of operation is reduced, but the manufacturing precision requirements are lowered
Solution Approach 1:
The system uses feedback from the front camera's gaze detection and probability map generation to automatically adjust the rear camera's capture settings. The detected eye direction and generated probability maps provide continuous feedback that enables automatic region of interest detection and setting adjustment, eliminating manual intervention
Solution Approach 2:
The imaging system performs self-service by automatically detecting the region of interest based on user gaze direction and autonomously adjusting capture settings without requiring manual user input. The system serves itself by using its own detection capabilities to control its own operation parameters
4Productivity
If automatic detection of region of interest is implemented, then the productivity of image capture is improved, but the measurement precision requirements increase
Solution Approach 1:
The front camera acts as an intermediary that captures user gaze information and generates probability maps, which then serve as input for the rear camera's region of interest detection. This intermediary mechanism enables automatic detection with high precision by using the front camera as a mediator between the user and the rear camera system
Data Source
AI summary
Disclosed is an image sensor, an image processing system including the image sensor, and an operating method of the image processing system, the image sensor including a first detection circuit configured to detect an eye region of a user from an information image obtained by capturing an image of the user, a first probability map generation circuit configured to generate a first probability map which corresponds to a direction in which the user's eyes gaze, based on the eye region, and a storage circuit configured to store the first probability map.