Multi-Frame Deinterlacing With Motion Sensing for Motion Blur
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video processing methods for interlaced videos suffer from poor de-interlacing effects, particularly in motion scenes, leading to issues like blurry details and image wire-drawing.
Innovation Solution
A video processing method using an image fusion model with a feature processing sub-model and a motion sensing sub-model to process at least three interlaced frames, enhancing both spatial and temporal features and motion sensing to produce high-quality target video frames.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-generated harmful factors
If traditional de-interlacing methods are used to process interlaced video, then the wire-drawing effect can be removed, but the image quality deteriorates with blurry details and wire-drawing in motion scenes
Solution Approach 1:
The patent introduces an image fusion model as an intermediary system between the interlaced video input and the final video output. This model includes a feature processing sub-model that extracts spatial-temporal features and a motion sensing sub-model that calculates motion vectors, serving as a mediator to transform the interlaced video into high-quality de-interlaced video without wire-drawing or blurring
Solution Approach 2:
The patent changes the processing parameters by using at least three interlaced frames instead of traditional pairwise processing, and by incorporating both spatial and temporal feature extraction along with motion vector calculation. This parameter change enables the system to achieve superior de-interlacing quality without the harmful effects of traditional methods
2Productivity
If traditional de-interlacing algorithms are applied, then processing speed is maintained, but detail loss occurs in motion scenes
Solution Approach 1:
The patent performs preliminary feature extraction and motion vector calculation on multiple interlaced frames before generating the final output. By pre-processing the spatial-temporal features and motion information from at least three frames, the system prepares the necessary data to reconstruct high-quality video frames without detail loss while maintaining efficient processing speed
Solution Approach 2:
The patent adds temporal dimension to the feature processing by using at least three interlaced frames and calculating motion vectors that capture temporal relationships. This dimensional expansion from traditional pairwise processing to multi-frame processing enables better detail recovery in motion scenes while maintaining productivity
Data Source
AI summary
Embodiments of the present disclosure provide a video processing method and apparatus, an electronic device, and a storage medium. The method includes: obtaining at least three interlaced frames to be processed, wherein each interlaced frame to be processed is determined based on two adjacent video frames to be processed; inputting the at least three interlaced frames to be processed to an image fusion model obtained by pre-training, to obtain at least two target video frames corresponding to the at least three interlaced frames to be processed, wherein the image fusion model includes a feature processing sub-model and a motion sensing sub-model; and determining a target video based on the at least two target video frames.