Low-Light Image Processing With Frame Interpolation for Smooth Video
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing photographing devices struggle with poor video fluency in low-light conditions due to limitations in ambient brightness and device photosensitivity.
Innovation Solution
An image processing method that adjusts frame rates based on ambient light levels, performing frame interpolation when light is low to increase the frame rate of collected video data, and using threshold-based frame dropping to stabilize frame rates as light conditions change.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If the sensor outputs a lower frame rate in low-light conditions to maintain photosensitivity, then the device can capture more light per frame, but the video fluency deteriorates
Solution Approach 1:
The patent introduces frame interpolation technology as an intermediary mechanism between the low frame rate captured in low-light conditions and the desired high frame rate output. The processing unit generates intermediate frames through interpolation algorithms, effectively mediating the contradiction by providing smooth video playback without requiring high capture frame rates, thus preserving light sensitivity while improving perceived fluency.
Solution Approach 2:
The patent dynamically changes the frame rate parameter based on ambient light conditions. In low-light scenarios, the sensor operates at a lower frame rate to maximize light capture, while the processing unit transforms this low frame rate input into a high frame rate output through frame interpolation. This parameter transformation resolves the contradiction by decoupling the capture frame rate from the output frame rate.
2Device complexity
If the device uses a single light brightness threshold for frame dropping, then the control logic is simple, but sudden changes in image brightness and noise occur when ambient light changes near the threshold
Solution Approach 1:
The patent segments the light brightness control into multiple thresholds (first threshold and second threshold) rather than using a single threshold. This segmentation creates distinct control zones: below the first threshold, frame dropping is enabled; between the first and second thresholds, the state is maintained; above the second threshold, frame dropping is disabled. This segmentation prevents sudden state transitions and the associated image quality fluctuations.
Solution Approach 2:
The patent implements a hysteresis mechanism using two thresholds that provides cushioning against sudden changes. When the ambient light brightness is between the first and second thresholds, the system maintains its previous state (either frame dropping enabled or disabled), preventing abrupt transitions. This beforehand cushioning ensures smooth transitions and stable image quality as light conditions change.
3Speed
If the device increases frame rate in well-lit conditions, then video fluency is improved, but the complexity of adaptive frame rate adjustment increases
Solution Approach 1:
The patent implements a feedback mechanism where the processing unit continuously monitors ambient light brightness and automatically adjusts whether to enable or disable frame dropping accordingly. This feedback loop simplifies the user experience while achieving adaptive frame rate optimization - the system self-adjusts based on real-time light conditions without requiring manual intervention or complex user configuration.
Data Source
AI summary
An image processing method includes an electronic device that outputs a first frame rate when the electronic device detects that ambient light brightness in a current environment is less than or equal to a first preset light brightness threshold; collects first video data based on the first frame rate, where the first video data includes a plurality of frames of images; and performs frame interpolation on the frames of images to obtain and store second video data, where a frame rate of the second video data is greater than the first frame rate.