Camera Scenario Recognition Continuity Under Device Shake
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing AI scenario recognition in electronic devices is prone to inaccurate or jumping results due to device shaking and object changes, leading to reduced user experience and increased computing overheads.
Innovation Solution
A method for determining a recommended scenario by assessing the continuity of the photographed scenario based on the camera's movement distance, geographical location, and posture changes, using preset conditions to ensure consistent scenario identification and reduce computing overheads.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If AI scenario recognition is performed based on image frames, then scenario recognition capability is provided, but device shaking causes inaccurate or jumping results
Solution Approach 1:
The patent introduces an optical image stabilization apparatus as an intermediary component between the camera and the image processing system. This apparatus compensates for device shaking by adjusting the optical path or sensor position, thereby isolating the AI scenario recognition system from the harmful effects of hand-held device movement and providing stable input images for accurate recognition.
Solution Approach 2:
The system performs preliminary stabilization of the captured images using optical image stabilization technology before the AI scenario recognition process. By pre-compensating for shaking effects, the system prevents the propagation of motion artifacts into the recognition algorithm, ensuring that the input images are stable and suitable for reliable scenario identification.
2Reliability
If stabilization apparatus is added to reduce camera jitter, then scenario recognition stability is improved, but device complexity and cost increase
Solution Approach 1:
The patent employs software-based image stabilization that creates a virtual copy or corrected version of the captured image frames. Instead of requiring complex hardware modifications, the system processes the original frames through algorithms that simulate the effect of physical stabilization, thereby achieving scenario recognition stability without significantly increasing device complexity or cost.
Solution Approach 2:
The system replaces or supplements mechanical/optical stabilization mechanisms with software-based image processing methods. By using computational algorithms to correct for camera jitter and stabilize images, the system achieves similar stabilization effects without the complexity and cost of additional mechanical stabilization apparatus.
3Ease of operation
If continuous scenario recognition is performed, then user experience is improved, but computing overheads increase
Solution Approach 1:
The system implements periodic scenario recognition at strategically selected time intervals rather than continuous processing. By recognizing scenarios at key moments (such as when the device is relatively stable or when scene changes are detected), the system maintains good user experience while significantly reducing the frequency of computationally intensive recognition operations and lowering overall computing overheads.
Solution Approach 2:
The system performs partial scenario recognition by focusing computational resources on processing only the most relevant or changed portions of the image stream. Instead of fully processing every single frame, the system selectively applies recognition algorithms to frames that are likely to contain meaningful scene changes, thereby reducing total computing overhead while maintaining effective scenario detection for user experience.
Data Source
AI summary
This application provides a method for determining a recommended scenario and an electronic device. The method includes: determining a recommended scenario of a current image based on a moving distance of a camera on an electronic device and a relationship between a current image and a previous image. When a slight shake occurs on the electronic device or the camera on the electronic device, or a photographed object/scenario does not change violently, continuity of a recommended scenario is preferably ensured, a probability of a jump of the recommended scenario is reduced, and user experience during image shooting is greatly improved.