Airborne Sensor Tracking Cluster Merging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current airborne tracking systems face challenges with track multiplicity, where multiple detections of the same object of interest lead to incorrect steering of sensors, resulting in loss of the correct reference to the object, especially when tracking boats on water due to false detections from whitecaps and wakes.
Innovation Solution
The system groups tracks associated with the object of interest into clusters and estimates the most likely location within these clusters, outputting a single track for the object, which is used by the automatic steering system to maintain the sensor aimed at the object, using image segmentation to classify pixels as either object or water class and associating detections with tracks based on bounding box overlap.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple detections are generated on the object of interest, then the detection coverage is improved, but track multiplicity occurs causing the tracking system to lose the correct reference
Solution Approach 1:
The patent merges multiple tracks that correspond to the same physical object into a single unified track. The track merging module combines detection results from multiple sources (optical sensor, radar, infrared) and consolidates them into one track representation, preventing track multiplicity while maintaining comprehensive detection coverage.
Solution Approach 2:
The patent introduces a track management module as an intermediary between detection and tracking components. This module acts as a mediator that receives multiple detections, resolves ambiguities, and generates a single authoritative track, thereby preventing track multiplicity while preserving detection reliability.
2Measurement precision
If the sensor is steered to keep the object in center of field of view, then tracking precision is improved, but the system cannot handle multiple tracks simultaneously
Solution Approach 1:
The patent extracts the track management function as a separate, dedicated module that handles all track-related operations independently. This extracted module manages track creation, maintenance, and merging without interfering with the sensor steering mechanism, simplifying the overall system architecture while maintaining precision.
Solution Approach 2:
The patent segments the tracking system into distinct functional modules: detection module, track management module, and sensor control module. Each module handles specific tasks independently, reducing complexity while maintaining high measurement precision through specialized processing in each segment.
3Difficulty of detecting and measuring
If false detections from whitecaps and wakes are included, then detection sensitivity is improved, but tracking reliability deteriorates due to incorrect object identification
Solution Approach 1:
The patent implements feedback mechanisms where track history and contextual information are continuously fed back into the detection process. The system uses past track data to evaluate current detections, suppressing false positives from whitecaps and wakes while maintaining sensitivity to genuine objects through adaptive threshold adjustment based on historical performance.
Solution Approach 2:
The patent performs preliminary evaluation of detections before they are converted into tracks. The system pre-processes detection results by comparing them against known object characteristics and historical data, filtering out false detections from whitecaps and wakes before they can contaminate the tracking system, thereby maintaining both sensitivity and reliability.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Airborne tracking systems use sensors to track objects of interest. In order to track the objects of interests, the sensors need to be steered such that the object is kept, ideally, in the center of the sensors field of view. Automatic steering of optical sensors requires the generation of a track on an object of interest. When tracking boats on the water, current approaches to image processing may generate multiple detections on the object of interest. Embodiments of the present disclosure solve the track multiplicity problem by grouping tracks associated with the object of interest into a cluster and by estimating a most likely location of the object within the cluster of tracks. Based on the estimated location, embodiments of the present disclosure outputs a single track for the object. The single track is used by an automatic steering system to maintain a sensor aimed at the object of interest.