Snapshot Image Quality in Video Recording via Frame Buffering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video recording technologies in mobile devices struggle to capture high-quality images during video recording due to delays between the user's capture operation and the actual image capture, resulting in suboptimal image quality.
Innovation Solution
A method involving buffering multiple frames of images in a first buffer queue and using a preset RAW domain image processing algorithm to select and enhance the image quality of the captured frame, which includes deep learning networks for image quality enhancement in the RAW domain.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the mobile phone captures images during video recording using conventional methods, then the capture operation can be performed, but the image quality is suboptimal due to delays between capture operation and actual image capture
Solution Approach 1:
The system pre-buffers multiple consecutive video frames (e.g., 5 frames) in memory during video recording before the user triggers the snapshot operation. When the user taps the snapshot button, the system immediately selects and processes the appropriate pre-buffered frame, eliminating the capture delay. This preliminary buffering of image data allows instant snapshot capture without waiting for frame acquisition.
Solution Approach 2:
The system creates a copy of multiple consecutive video frames and stores them in a buffer queue during video recording. When snapshot is triggered, it retrieves a copy from the buffer rather than capturing a new frame, ensuring the captured image corresponds to the moment the user intended while avoiding processing delays. The buffer contains redundant frame copies for immediate access.
2Measurement precision
If multiple frames are buffered to enable frame selection for optimal quality, then image quality can be improved, but the device complexity increases
Solution Approach 1:
The system evaluates multiple pre-buffered frames using quality metrics (such as focus detection, exposure levels, and noise assessment) and automatically selects the frame with the optimal quality for snapshot capture. This feedback mechanism ensures high image quality by intelligently choosing the best frame from the buffer based on real-time quality assessment, rather than simply using the first or last frame.
Solution Approach 2:
The buffer management system operates autonomously during video recording, continuously pre-buffering frames and preparing them for potential snapshot capture without requiring user intervention. When snapshot is triggered, the system self-selects the appropriate frame from the buffer based on quality metrics and timestamp matching, performing the selection and processing automatically without additional user actions.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
This application relates to the field of photographing technologies, and provides a method for capturing an image during video recording and an electronic device, so that an image can be captured during video recording and image quality of a captured image is improved. The method includes: collecting, by an electronic device, a first image in response to a first operation, and displaying a first screen, where the first screen includes a snapshot shutter and a preview image obtained based on the first image; buffering, in a first buffer queue, the first image collected by a camera, where n frames of first images collected by the camera are buffered in the first buffer queue; selecting a second image from the n frames of first images based on additional information of the first images in response to a second operation on the snapshot shutter, where metadata includes contrast of the first image and an angular velocity at which the camera collects the first image; and running a preset RAW domain image processing algorithm, by using m frames of first images in the n frames of first images as an input, to obtain a third image, and encoding the third image to generate a captured photo, where the m frames of first images include the second image, and the preset RAW domain image processing algorithm has a function of improving image quality.