3-axis OIS for Super-resolution Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current super-resolution imaging techniques in handheld mobile devices are limited by 2-axis optical image stabilization (OIS) systems that fail to correct for camera roll, leading to motion blur and precision errors in subpixel shifts, which hinder accurate image registration and resolution enhancement.
Innovation Solution
Implementing a 3-axis OIS system that stabilizes images in three degrees of freedom (pitch, yaw, and roll) using a MEMS actuator to move the image sensor or lens, enabling subpixel precision and compensating for camera movements, and leveraging natural hand jitter to induce subpixel shifts without relying on actuator commands.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If 2-axis OIS is used to stabilize images during super-resolution imaging, then motion blur is reduced, but roll correction is insufficient leading to precision errors in subpixel shifts
Solution Approach 1:
The patent transitions from 2-axis OIS to 3-axis OIS by adding the roll axis (third degree of freedom) to the existing pitch and yaw stabilization. This dimensional expansion enables comprehensive correction of all camera movements including roll, thereby eliminating the precision errors in subpixel shifts that occurred with 2-axis systems while maintaining motion blur reduction capabilities
2Measurement precision
If actuator commands are used to induce subpixel shifts, then controlled image sampling is achieved, but precision is limited by actuator accuracy
Solution Approach 1:
The patent leverages the natural hand jitter of the user holding the device as the source of subpixel shifts, eliminating the need for complex actuator control systems. The 3-axis OIS system captures these natural movements and uses them for super-resolution imaging, thereby achieving high subpixel shift accuracy without additional mechanical complexity
Solution Approach 2:
The patent converts the harmful effect of hand jitter (which causes motion blur) into a beneficial source of subpixel shifts for super-resolution imaging. By using 3-axis OIS to stabilize the image while capturing natural hand movements, the system transforms what was previously a source of error into the mechanism for achieving subpixel precision
3Area of moving object
If digital zooming is applied to achieve higher resolution, then area of interest is magnified, but resolution is significantly less than physical pixel resolution
Solution Approach 1:
The patent captures multiple images continuously at different subpixel positions (using natural hand jitter and 3-axis OIS stabilization) and combines them through super-resolution processing. This continuous sampling at subpixel intervals allows digital zooming to achieve resolution that exceeds the physical pixel resolution of the sensor, maintaining high resolution while magnifying the area of interest
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach eliminates motion blur and achieves subpixel precision in image registration, significantly enhancing image resolution beyond the physical limits of the camera, allowing for digital zooming beyond the device's native resolution.
Implementation Method 1
The OIS system includes a MEMS actuator that moves the image sensor in the three degrees of freedom
Data Source
AI summary
Systems and methods are provided for super-resolution imaging using 3-axis OIS. A super-resolution image may be created by enabling optical image stabilization (OIS) in three axes using an OIS system on a camera of an image capturing device; capturing an image of a scene using an image sensor of the camera; shifting the image on the image sensor by a predetermined subpixel amount; capturing the subpixel shifted image; and constructing a super-resolution image of the scene using the image and the subpixel shifted image. In one particular implementation, a position sensor may measure a positional drift of the image sensor after capturing the image. Using this measured positional drift, a time sufficient to shift the image sensor by a predetermined subpixel amount may be determined. The OIS may subsequently be disabled in one or two axes for the determined time.