Digital Zoom Video Stabilization via Sensor-Driven Shake Compensation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital image stabilization methods for video capturing devices, such as smartphones, are inadequate in compensating for hand tremors and subject movement, leading to shaky digital zoom videos.
Innovation Solution
The method involves determining shaking compensation areas within the digital zoom video frames using sensors and motion information to adjust and stabilize the video, either by cropping a partial area or using the entire frame, thereby compensating for hand tremors and subject movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If digital zoom is applied to capture distant subjects, then the field of view is narrowed to focus on specific details, but hand tremors and subject movement cause increased video shakiness
Solution Approach 1:
The video stabilization process is segmented into distinct stages: shake detection from sensor data, shake amount calculation, and frame composition adjustment. This segmentation allows each stage to be optimized independently, enabling effective stabilization even at high digital zoom levels where traditional methods fail
Solution Approach 2:
Sensor data (accelerometer and gyroscopic information) serves as an intermediary to detect and quantify hand tremors and device movement. This intermediary measurement enables the system to compensate for shaking without directly interfering with the digital zoom function, resolving the contradiction between zoom magnification and video stability
2Stability of the object's composition
If shake compensation area is expanded to entire frame, then more area is available for stabilization adjustment, but processing complexity and computational load increase
Solution Approach 1:
The patent applies different processing strategies to different regions of the frame. The shake compensation is primarily focused on the subject area rather than uniformly processing the entire frame, reducing computational complexity while maintaining stabilization effectiveness. This local quality approach allows selective application of stabilization algorithms
3Adaptability or versatility
If multiple stabilization methods are provided (crop-based and full-frame), then user can select based on preferences, but device complexity and memory requirements increase
Solution Approach 1:
The system dynamically adjusts stabilization parameters and compensation areas based on detected shake magnitude and direction. Rather than providing completely separate processing pipelines for different methods, the system dynamically modulates a unified stabilization process, reducing memory requirements while maintaining multiple operational modes
Solution Approach 2:
Different stabilization methods are achieved by changing parameters within a unified framework: the shake compensation area can be adjusted between subject-region-only and full-frame modes, and the degree of compensation can be modulated based on shake severity. This parameter-based approach reduces complexity compared to implementing entirely separate processing systems
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 effectively stabilizes digital zoom videos by adjusting the video frames to counteract hand tremors and subject movement, resulting in a smoother playback experience.
Implementation Method 1
a sensor for detecting the shake of the device itself by the movement of the device
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
A method of stabilizing a video in an electronic device is disclosed. The method includes determining shaking compensation areas corresponding to a respective plurality of frames acquired through an image sensor in entire areas of the respective plurality of frames based on shaking information of the plurality of frames and information on the magnification, generating output frames corresponding to the respective plurality of frames by enlarging the shaking compensation areas at the magnification, and providing a video including the output frames.