Scene Detection for Periodic Video Motion Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video frame rate up-conversion techniques fail to accurately detect motion vectors in periodic scenes, leading to image quality issues like broken lines and blur due to deviations in motion estimation.
Innovation Solution
A scene detection method that calculates match errors between central and surrounding blocks within a search range, setting thresholds based on prediction depths to determine periodicity, allowing for accurate motion estimation and improved image quality by aligning with the periodic pattern of the scene.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If simple frame repetition or averaging adjacent frames is used to increase frame rate, then the frame rate is improved, but the video jitter and blur effect cannot be solved and image quality deteriorates
Solution Approach 1:
The patent performs scene detection and motion vector estimation before frame rate conversion. By preliminarily identifying periodic scenes and calculating accurate motion vectors, the system prepares the necessary information to guide subsequent frame interpolation, preventing jitter and blur while maintaining high frame rate output
Solution Approach 2:
The patent uses scene detection results and motion vector information as feedback to guide the frame rate conversion process. The system continuously monitors match errors between blocks and adjusts motion estimation based on periodic pattern detection, creating a closed-loop system that optimizes image quality while maintaining high frame rate
2Speed
If motion estimation is performed without scene detection, then the processing speed is maintained, but the accuracy of motion vector calculation deteriorates in periodic scenes leading to broken lines and blur
Solution Approach 1:
The patent divides the image into blocks and systematically compares each block with surrounding blocks to detect periodic patterns. By segmenting the image processing into block-based match error calculation and periodicity detection, the system maintains processing efficiency while achieving high accuracy motion vector estimation in periodic scenes
Solution Approach 2:
The patent changes the parameter of match error thresholding based on block position and surrounding block characteristics. By dynamically adjusting threshold values and search ranges according to the detected periodic pattern, the system optimizes motion vector accuracy without significantly increasing processing time
Data Source
AI summary
The present disclosure relates to a scene detection method, a chip, an electronic device, and a storage medium, the present disclosure detects periodic scenes, which helps to improve the accuracy of motion estimation.