Image Tracking via Background Frame Segmentation and Similarity Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing background generation methods fail to accurately track moving objects in videos, resulting in imprecise object tracking due to difficulties in generating a complete background.
Innovation Solution
An electronic device with modules for image capturing, processing, and display units that capture and process video frames to set detecting regions, calculate background frames, and track moving targets by determining similarities between reference and detecting regions, adjusting frames to fill gaps, and using grayscale conversion for accurate positioning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing background generation methods are used, then the processing is simple, but the tracking precision deteriorates
Solution Approach 1:
The patent segments the video processing into distinct stages: background frame generation, detecting region division, and similarity calculation. By dividing the complex tracking task into manageable segments, the system achieves precise tracking while maintaining organized processing flow. Each segment handles specific aspects of the problem independently.
Solution Approach 2:
The patent performs preliminary actions by pre-generating stable background frames and pre-dividing detecting regions before actual tracking. This preliminary preparation ensures that when tracking occurs, the computational foundation is already established, improving both precision and efficiency without adding complexity during real-time operation.
Solution Approach 3:
The patent introduces grayscale conversion as an additional dimension in the tracking process. By converting color images to grayscale and calculating similarity based on grayscale values rather than full color information, the system simplifies calculations while maintaining tracking precision, effectively adding a dimensional transformation to resolve the contradiction.
2Productivity
If background generation is simplified, then the processing speed improves, but the completeness of background generation deteriorates
Solution Approach 1:
The patent applies partial action by focusing background generation on stable regions and using simplified methods for dynamic regions. Rather than applying complex generation methods uniformly across the entire frame, the system selectively applies appropriate methods to different regions, improving overall processing speed while maintaining completeness where needed.
Solution Approach 2:
The patent implements local quality by allowing different background generation approaches for different regions of the frame. Stable regions receive more complete generation treatment while dynamic regions use faster methods. This regional differentiation ensures processing speed is optimized without sacrificing background completeness in critical areas.
3Measurement precision
If detecting regions are increased, then the tracking accuracy improves, but the computational complexity increases
Solution Approach 1:
The patent uses partial action by calculating similarity only for divided detecting regions rather than the entire frame. This selective approach focuses computational resources on relevant areas, improving tracking accuracy through targeted analysis while reducing overall computational load by excluding unnecessary regions from processing.
Solution Approach 2:
The patent segments the detecting regions into multiple divided regions, allowing independent similarity calculation for each segment. This segmentation improves tracking accuracy by enabling fine-grained analysis of different areas while managing computational load through modular processing of smaller regions rather than one comprehensive calculation.
Data Source
AI summary
An electronic device able to receive a target frame for tracking a moving target which is included in a search reference sets at least one first detecting region in the target frame. The processor determines similarities between the search reference and the at least one first detecting regions. Then, the processor determines a specific first detecting region in the target frame and determines an actual position of the moving target based on the specific first detecting region.