Radar-Aided Video Tracking for Redundant Track Removal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video tracking systems fail to effectively identify and eliminate redundant video tracks of related objects, leading to unnecessary processing and potential false alarms.
Innovation Solution
A method utilizing radar tracks to identify redundant video tracks by comparing video tracks with radar tracks based on position and classification similarity, where radar tracks indicate related objects, allowing for the deletion or ignoring of redundant video tracks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If video tracking detects all moving objects including related objects, then detection completeness is improved, but processing load and false alarms increase due to redundant tracks
Solution Approach 1:
The patent introduces radar tracking as an intermediary system to verify video tracks. Radar tracks serve as a mediator to distinguish between main objects and related objects, allowing the system to maintain high detection completeness while reducing processing load by filtering out redundant video tracks through cross-validation with radar data
Solution Approach 2:
The system implements feedback by using radar tracking results to validate and filter video tracks. The radar track information feeds back into the video tracking system to identify and eliminate redundant tracks, creating a closed-loop verification process that reduces false alarms and processing load while maintaining detection accuracy
2Measurement precision
If video tracking creates separate tracks for related objects, then object classification accuracy is improved, but system complexity increases due to redundant track management
Solution Approach 1:
Radar tracking serves as an intermediary verification layer that simplifies track management. By using radar tracks as a reference, the system can automatically determine which video tracks are redundant without complex manual classification rules, reducing system complexity while maintaining classification accuracy
Solution Approach 2:
The system changes the parameter basis for track validation by incorporating radar-specific parameters (position, velocity, classification) into the verification process. This multi-parameter approach allows automatic differentiation between main objects and related objects, reducing management complexity while preserving classification precision
3Reliability
If radar and video tracks are cross-validated, then false alarm reduction is improved, but computational requirements increase
Solution Approach 1:
The system extracts only the essential verification parameters from radar tracks (position, velocity, classification) that are necessary for validating video tracks. By taking out only the critical parameters rather than processing complete radar datasets, the system achieves false alarm reduction while minimizing additional computational energy requirements
Data Source
Figure 1a
Figure 1b
Figure 2
AI summary
A device 200, a non-transitory computer-readable storage medium, and a computer implemented method 100 are provided for identifying redundant video tracks. Video tracks and a radar track are obtained S110, S120. The video tracks comprise positions and a classification in relation to respective objects detected in a sequence of video image frames captured of a scene during a time period and the radar track comprises positions and a classification in relation to of an object detected in radar data captured for at least a part of the scene during the time period. An indication is obtained 5130 for each video track of the video tracks whether the radar track and the video track of the video tracks correspond based on a similarity according to a similarity measure. A second video track is identified S140 as redundant on condition that the radar track and a first video track correspond, the classifications associated with the radar track and the first video track correspond, the radar track and the second video track correspond, and the classifications associated with the radar track and the second video track do not correspond.