Target Tracking Box Matching for Occlusion Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing target tracking algorithms fail to accurately track targets when they are shielded by other objects, such as people or vehicles, leading to inaccurate tracking results.

Innovation Solution

A method that involves matching target tracking boxes with candidate boxes to determine shielding relationships, using Intersection over Union (IoU) and Munkres Assignment Algorithm to identify best matching boxes, and restarting tracking when shielding occurs, employing Person Re-identification (ReID) networks to recover tracking.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional target tracking algorithms are used, then the tracking process is simple, but the tracking accuracy deteriorates when the target is shielded by other objects

Engineering Contradiction:

Improvetracking accuracyVSAvoidtracking algorithm complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the tracking problem into multiple independent tasks: candidate box generation, matching with tracking box, shielding relationship determination, and restart decision-making. This segmentation allows each component to be optimized independently while improving overall tracking accuracy under shielding conditions.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces intermediary elements including candidate boxes as intermediaries between the detector and tracker, and a matching relationship determination module as an intermediary to analyze spatial relationships. These intermediaries enable the system to detect shielding conditions and make informed decisions about restarting tracking.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Reliability

If the target tracking algorithm continuously tracks without restarting, then the tracking process is continuous, but false detections increase when shielding occurs

Engineering Contradiction:

Improvetracking reliabilityVSAvoidtracking interruption time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent implements feedback mechanisms where the matching relationship determination module continuously monitors the spatial relationships between candidate boxes and tracking box, providing feedback about shielding conditions. This feedback loop enables the system to detect when shielding occurs and trigger appropriate restart actions to maintain tracking reliability.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent introduces dynamic decision-making where the system adapts its tracking behavior based on real-time shielding detection. When shielding is detected through matching relationship analysis, the system dynamically restarts tracking to recover from false detections, balancing continuity with reliability.

Inventive Principle:

Principle #15Dynamics

3Measurement precision

If matching analysis is performed on all candidate boxes, then the shielding detection is comprehensive, but the computational complexity increases

Engineering Contradiction:

Improveshielding detection accuracyVSAvoidcomputational power consumption

Core Design Contradiction:

Measurement precisionVSPower

Solution Approach 1:

The patent applies partial action by performing matching analysis selectively rather than on all candidate boxes uniformly. The system focuses computational resources on analyzing matching relationships that are most likely to indicate shielding conditions, achieving comprehensive shielding detection with reduced computational overhead.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS12430772B2Target tracking method, computer-readable storage medium, and computer device

Publication Date: 2025.09.30 ARASHI VISION INC

AI summary

A target tracking method, a computer-readable storage medium, and a computer device. The method comprises: matching a target tracking box with target candidate boxes to determine a target candidate box in best matching with the target tracking box; matching one or more remaining target candidate boxes, except for the best matching target candidate box, with a second target candidate box detected previously to determine a corresponding matching relationship; according to the best matching target candidate box and the corresponding matching relationship, obtaining distances and overlapping relationships respectively between the best matching target candidate box and the one or more remaining target candidate boxes and between the best matching target candidate box and the second target candidate box, so as to determine a shielding relationship between a target and other objects in the current video frame; and determining, according to the shielding relationship, whether to restart target tracking.