Adaptive Frame Rate Conversion Using Region-Based Motion Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing frame rate conversion methods, particularly motion compensated (MC) based interpolation, face challenges in scenes with fast-moving objects, occlusions, and areas with little detail, resulting in visible artifacts and inefficient use of hardware capacity.
Innovation Solution
The proposed solution involves building candidate lists using motion vector flow and regional motion vectors, employing small range search, and adaptively choosing winner candidates with spatial and temporal biases, along with post-motion estimation correction and adaptive motion compensation to handle occlusions and logo detection, ensuring reliable interpolation and minimizing artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If motion estimation is performed by selecting winner motion vectors from candidate vectors in prior art MEMC methods, then the method works well for scenes with consistent motions, but it results in lots of visible artifacts for difficult cases such as boundaries of fast moving objects, occlusion scenes, and areas with little details
Solution Approach 1:
The patent divides the image into multiple regions and performs motion estimation separately for each region. By segmenting the image into blocks and processing motion vectors region-by-region, the method can handle different motion characteristics in different areas, reducing artifacts at boundaries and in occlusion regions while maintaining accuracy for consistent motion areas.
Solution Approach 2:
The patent implements adaptive motion estimation that dynamically adjusts the search range and candidate selection based on local motion characteristics. The system detects occlusion scenes and fast-moving object boundaries, then modifies the motion compensation strategy accordingly, making the algorithm flexible rather than rigid to handle diverse scene conditions.
2Productivity
If traditional MEMC algorithms are used for frame rate conversion, then hardware capacity can be utilized, but visual quality deteriorates in challenging scenes with occlusions and fast moving objects
Solution Approach 1:
The patent applies different motion estimation and compensation strategies to different regions of the image based on their specific characteristics. In regions with occlusions or fast motion, the algorithm uses adjusted search ranges and candidate selection criteria, while in stable regions it uses standard MEMC. This localized approach maintains visual quality without sacrificing overall processing efficiency.
Solution Approach 2:
The system dynamically changes key parameters of the motion estimation algorithm based on scene analysis. When occlusion or fast motion is detected, the search range, candidate vector selection, and compensation weights are adjusted. This parameter adaptation allows the system to maintain high visual quality across diverse scene types while preserving hardware utilization efficiency.
Data Source
AI summary
The disclosure is regarding an apparatus and a method for frame rate conversion. The frame rate conversion apparatus, for receiving at least two input frames to generate at lease one output frame, comprises a motion estimation module for generating a candidate list comprising regional motion vectors and temporal motion vectors from the at least two input frames and choosing a winner motion vector from the candidate list, a motion compensation module for generating the at lease one output frame according to the winner motion vector, a frame feature extraction module for generating fade-in/out information or a logo detection information from the at least two input frames, and an adaptive MEMC control module for deciding a weighting of the winner motion vector according to the fade-in/out information or the logo detection information.