Video Object Boundary Detection via Temporal Patch Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing edge detection algorithms in computer vision struggle to differentiate between object boundaries and texture edges due to the reliance on edge strength, which does not distinguish between the two types effectively, and lack a clear definition of what constitutes an object contour versus a texture edge.
Innovation Solution
The method involves tracking image patches over a temporal window in video frames to determine object boundary probabilities by identifying patch matches and using metrics such as color and gradient information, error mean, k-means clustering, bimodality, histogram comparison, cross-correlation, or supervised learning to calculate probability maps that distinguish between object boundaries and texture edges.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Difficulty of detecting and measuring
If edge strength is used to detect object boundaries, then edge detection can be performed, but the ability to distinguish between object boundaries and texture edges deteriorates
Solution Approach 1:
The patent transitions from analyzing single-frame edge strength to analyzing temporal consistency across multiple video frames. By adding the time dimension and tracking patch matches across frames, the system can distinguish object boundaries (which move with the object) from texture edges (which remain stationary relative to the background), thereby resolving the contradiction between detecting edges and differentiating their types.
2Measurement precision
If a clear definition of object contour versus texture edge is established, then edge classification improves, but the complexity of the detection system increases
Solution Approach 1:
The system uses temporal feedback by tracking patches across multiple video frames. The consistency of patch matches over time provides feedback that enables automatic differentiation between object boundaries and texture edges without requiring complex manual definitions, thus improving classification accuracy while maintaining reasonable system complexity.
Data Source
AI summary
Methods and systems described herein detect object boundaries of videos. A window around the pixel may be followed in adjacent image frames of the image frame to determine object boundaries. Inconsistencies in image patches over a temporal window are detected, and each pixel of the image frame of a video is assigned an object boundary probability. The pixel may belong to a texture edge if the window content does not change throughout the adjacent image frames, or the pixel may belong to an object boundary if the window content changes. A probability value indicating the likelihood of the pixel belonging to an object boundary is determined based on the window content change and is assigned to the corresponding pixel.