Eye Tracking Camera Tilting for ROI Depth Sensing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional depth sensors for small devices like augmented reality glasses consume high power and require extensive calculations to obtain depth information for the entire space, making it difficult to mount them on small devices and reducing accuracy and response speed.
Innovation Solution
An electronic device equipped with an eye tracking sensor, a camera, and a tilting unit that adjusts the camera direction to obtain depth information only for the region of interest (ROI) by capturing images from different angles, using the eye tracking sensor to determine the user's gaze point and calculate depth information based on these images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If depth sensing is performed on the full range of spaces using a depth sensor, then depth information of the entire space is obtained, but power consumption increases and calculation amount increases
Solution Approach 1:
The patent applies local quality by shifting from full-space depth sensing to ROI-specific depth sensing. The system uses an eye tracking sensor to identify the user's gaze point and determines a region of interest (ROI) around this point. Depth sensing is then performed only within this localized ROI rather than the entire measurable space, reducing power consumption while maintaining depth information quality for the most relevant area.
Solution Approach 2:
The patent segments the measurable space into multiple regions, with the full space divided into a region of interest (ROI) and other areas. By segmenting the depth sensing task to focus only on the ROI portion that contains the user's gaze point, the system reduces the overall calculation amount and power consumption while still providing comprehensive depth information where needed.
2Measurement precision
If depth sensing is performed on the full range of spaces, then complete depth information is obtained, but device size increases making it difficult to mount on small devices
Solution Approach 1:
The patent reduces device size by implementing local quality - performing depth sensing only on the ROI rather than the full measurable space. This localized approach allows the depth sensor and associated processing components to be miniaturized, making it feasible to mount the depth sensing system on small devices like augmented reality glasses while still providing complete depth information for the user's field of view.
Solution Approach 2:
By segmenting the depth sensing task to focus only on the ROI containing the user's gaze point, the patent reduces the computational burden and hardware requirements. This segmentation enables the use of smaller, more power-efficient depth sensors and processing units that can be integrated into compact device form factors.
3Measurement precision
If depth sensing is performed on the full range of spaces, then all spatial depth information is captured, but calculation amount and processing time increase
Solution Approach 1:
The patent improves calculation speed by applying local quality - concentrating depth sensing resources on the ROI rather than processing the entire measurable space. The system uses the eye tracking sensor to identify the user's gaze point, determines the ROI, and performs depth sensing only within this focused region, significantly reducing the calculation amount while maintaining spatial coverage where it matters most for user interaction.
Solution Approach 2:
The patent segments the depth sensing process into identifying the gaze point, determining the ROI, and performing depth sensing only within the ROI. This segmentation of the calculation process reduces the overall computational load and processing time compared to performing depth sensing on the full range of spaces, thereby improving productivity and response speed.
Data Source
AI summary
An electronic device including an eye tracking sensor configured to detect viewing direction of eye of a user; a camera; a tilting unit configured to adjust a direction of the camera; a memory; and a processor configured to obtain information about a gaze point of the user; determine a region of interest (ROI) based on the information about the gaze point; obtain two or more images including the ROI at different tilt angles; and determine depth information of the ROI by using the obtained two or more images.