Partial Motion Blur Detection via Feature Point Speed Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing blur detection algorithms fail to identify video frames with partial motion blur, leading to false positives in selecting high-quality stills, as they only detect frames that are completely or nearly completely blurry, missing frames where only part of the content is blurry.
Innovation Solution
The method involves identifying feature points in a video clip, calculating their speeds, determining a collective speed for each frame, and using a selection factor based on this speed to select frames that are relatively free of motion blur, even if blur occurs in only a part of the image, by comparing the selection factor to a threshold.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing blur detection algorithms are used to identify clear frames, then completely blurry frames are detected and excluded, but frames with partial motion blur are missed and included as false positives
Solution Approach 1:
The patent segments the video frame into multiple regions and tracks feature points across these regions to detect motion blur locally. By dividing the frame analysis into discrete feature point tracking, the system can identify partial motion blur in specific areas without discarding the entire frame, thus resolving the contradiction between detecting complete blur and preserving frames with partial blur.
Solution Approach 2:
The patent applies local quality analysis by evaluating different regions of the frame independently through feature point tracking. Instead of treating the entire frame uniformly, it assesses the motion blur characteristics of specific local areas, allowing frames with partial motion blur to be correctly identified and handled based on their local quality characteristics.
2Measurement precision
If manual frame selection is performed, then high-quality stills can be identified, but the process is tedious and infeasible for large video datasets
Solution Approach 1:
The patent implements self-service automation where the system automatically performs frame selection based on computed motion blur metrics. By enabling the system to self-evaluate and self-select frames without human intervention, it eliminates the time-consuming manual selection process while maintaining accurate quality assessment through automated feature point tracking and motion blur detection.
Solution Approach 2:
The patent transforms the manual selection process into an automated parameter-based decision system. By computing objective parameters such as motion blur magnitude and distribution, the system can automatically rank and select frames based on quantifiable metrics, replacing subjective manual evaluation with objective automated assessment that scales efficiently with dataset size.
3Measurement precision
If feature point speed calculation is performed for all frames, then motion blur detection accuracy improves, but processing time and computational complexity increase
Solution Approach 1:
The patent applies partial action by calculating feature point speeds only for frames and regions where motion blur detection is necessary, rather than uniformly processing all frames. By selectively applying the computationally intensive feature tracking only where needed based on preliminary analysis or frame characteristics, the system maintains high detection accuracy while reducing overall processing complexity and time requirements.
Data Source
AI summary
An automated motion-blur detection process can detect frames in digital videos where only a part of the frame exhibits motion blur. Certain embodiments programmatically identify a plurality of feature points within a video clip, and calculate a speed of each feature point within the video clip. A collective speed of the plurality of feature points is determined based on the speed of each feature point. A selection factor is compared to a selection threshold for each video frame. The selection factor is based at least in part on the collective speed of the plurality of feature points. Based on this comparison, at least one video frame from within the video clip is selected. In some aspects, the selected video frame is relatively free of motion blur, even motion blur that occurs in only a part of the image.