Motion Vector Correction for Video Interpolation Artifacts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Video signal processing apparatuses face challenges in reducing the feeling of visual strangeness caused by false detection of motion vectors, especially when the detected motion vector is relatively large, leading to enhanced visual strangeness in interpolated images.
Innovation Solution
A motion vector correction device and method that includes a motion vector detector and corrector, which detect and correct motion vectors using pixel data from real frames to generate interpolated frames, with a motion vector corrector that suppresses and converts motion vector magnitudes based on predefined thresholds and state determination values to minimize false detection impacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If motion vector detection is performed using pixel data within limited ranges in successive frames, then frame rate conversion can be carried out, but false detection of motion vector occurs leading to visual strangeness in interpolated images
Solution Approach 1:
The patent applies preliminary action by detecting motion vectors not only in successive frames but also in frames prior to the current frame and future frames. This advance detection across multiple time points allows the system to identify and correct false motion vector detections before they affect the interpolated image quality, thereby maintaining both high frame rate conversion capability and accurate motion vector detection.
2Measurement precision
If motion vector magnitude is large, then more detailed motion information is captured, but false detection causes enhanced visual strangeness in interpolated images
Solution Approach 1:
The patent implements feedback by determining whether a detected motion vector is abnormal based on its magnitude and temporal characteristics. When a motion vector's magnitude exceeds a threshold or shows inconsistent patterns across multiple frames, the system identifies it as abnormal and corrects it. This feedback mechanism allows the system to maintain high measurement precision for motion vectors while preventing visual strangeness by correcting false detections.
Solution Approach 2:
The patent applies parameter changes by dynamically adjusting the threshold for motion vector magnitude based on the temporal variance of motion vectors. When motion vectors show high temporal variance indicating potential false detection, the threshold is adjusted to be more stringent. This adaptive parameter adjustment allows the system to maintain high measurement precision while reducing visual strangeness caused by false detections.
3Object-affected harmful factors
If motion vector correction is applied to reduce visual strangeness, then interpolated image quality improves, but device complexity increases
Solution Approach 1:
The patent applies universality by implementing a motion vector correction mechanism that works across multiple frames and applies the same correction logic consistently. The system uses a unified approach where abnormal motion vectors are detected and corrected based on temporal variance and magnitude thresholds, making the correction mechanism universally applicable to different video sequences without requiring complex frame-specific processing. This reduces overall device complexity while maintaining improved interpolated image quality.
Data Source
AI summary
A motion vector detector detects, using pixel data in at least two real frames in an input video signal, a motion vector MV1 necessary for generating interpolated pixel data forming an interpolated frame to be inserted between the two real frames. A motion vector corrector corrects the motion vector MV1 to decrease the magnitude of motion vector MV1 when the magnitude of motion vector MV1 exceeds a predetermined threshold, and outputs it as a motion vector MV3.