Motion Vector Calibration for Video Foreground Background Separation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing motion compensation methods inaccurately assign motion vectors at the edges of objects or between foreground and background in videos, leading to poor quality high frame rate videos due to mismatching macroblocks.
Innovation Solution
A video processing method that divides frames into blocks, generates motion vectors, calculates global motion vectors, computes sums of absolute pixel differences, matches these with predefined models to label blocks as foreground or background, and modifies motion vectors accordingly to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If motion compensation divides video into small macroblocks and calculates motion vector for each macroblock, then processing speed and local motion detection are improved, but motion vector accuracy at object edges and foreground-background junctions deteriorates
Solution Approach 1:
The patent segments the video frame into multiple regions based on motion characteristics. It divides the frame into foreground and background regions using motion vector analysis, then applies different motion compensation strategies to each region. This segmentation allows the system to maintain high processing speed while improving accuracy at critical boundaries by treating different regions differently.
Solution Approach 2:
The patent applies local quality by using different motion vector calculation methods for different regions. For background regions, it uses standard macroblock-based motion estimation. For foreground regions and their boundaries, it employs refined motion vector calculation that considers edge characteristics and junction points, thereby improving local motion vector accuracy without compromising overall processing efficiency.
2Device complexity
If standard macroblock motion estimation is used, then computational complexity is reduced, but motion vector mismatch at object edges and foreground-background junctions increases
Solution Approach 1:
The patent performs preliminary action by first identifying foreground and background regions before final motion compensation. It calculates initial motion vectors for all macroblocks, then identifies regions with potential motion vector mismatches (such as object edges and foreground-background junctions). This preliminary identification allows the system to apply refined correction only where needed, reducing overall computational complexity while improving reliability at critical locations.
Solution Approach 2:
The patent introduces an intermediary step of motion vector calibration. After initial motion estimation, it calculates a calibration factor based on the difference between reference frame and current frame characteristics, then applies this calibration to adjust motion vectors in problematic regions. This intermediary calibration process resolves motion vector mismatches without requiring complete recalculation, thus maintaining low computational complexity while improving matching accuracy.
Data Source
AI summary
A video processing method includes dividing a current frame into a plurality of blocks, generating a motion vector of each block of the plurality of blocks of the current frame according to the each block of the current frame and a corresponding block of a previous frame, generating a global motion vector according to a plurality of motion vectors of the current frame, generating a sum of absolute differences of pixels of each block of the current frame according to the global motion vector, generating a region with a set of blocks of the current frame, matching a distribution of the sum of absolute differences of pixels of the region with a plurality of models, identifying a best matching model, and labeling each block in the region in the current frame with a label of either a foreground block or a background block according to the best matching model.