Object Tracking Apparatus with Associated Tracking for Overlapping Targets
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object tracking systems using video data struggle with flexibility in tracking multiple objects, particularly when one object enters or rides on another, making direct tracking challenging.
Innovation Solution
An object tracking apparatus and method that includes a tracking processing unit and a detection unit. The detection unit identifies when a predetermined condition is met for a first object and a second object, allowing the tracking processing unit to perform tracking processing of the second object in association with the first object.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If direct tracking processing is performed on all objects, then tracking precision is maintained, but device complexity and processing load increase when objects interact (one object enters or rides on another)
Solution Approach 1:
The tracking system dynamically adjusts its processing mode based on detected conditions. When objects are determined to be separate, independent tracking is performed. When one object enters or rides on another (overlap detected), the system switches to associated tracking where the second object's tracking results are used to represent the first object's position, thereby adapting processing complexity to actual tracking needs
Solution Approach 2:
The system introduces a detection unit that acts as an intermediary between video data and tracking processing. This detection unit determines whether objects are overlapping or separate, and based on this determination, controls whether associated tracking or independent tracking should be performed, mediating the tracking process according to object relationships
2Productivity
If independent tracking is performed for each object, then processing speed is maintained, but adaptability decreases when objects have spatial relationships (one object enters or rides on another)
Solution Approach 1:
The tracking approach dynamically changes based on object spatial relationships. When objects are separate, fast independent tracking is used. When objects overlap (one enters or rides on another), the system transitions to associated tracking that leverages the spatial relationship, thereby maintaining processing efficiency while adapting to different object interaction scenarios
Solution Approach 2:
The tracking processing unit is designed to perform multiple functions: it can track objects independently when they are separate, and it can perform associated tracking when objects have spatial relationships. This multi-functionality allows the system to handle various tracking scenarios (separate objects, overlapping objects, objects with spatial relationships) using a single unified system
3Adaptability or versatility
If associated tracking is performed for objects with spatial relationships, then tracking flexibility improves, but measurement precision may decrease when direct tracking is feasible
Solution Approach 1:
The system dynamically selects the appropriate tracking mode based on real-time detection of object relationships. When objects are determined to be separate (no overlap), independent tracking with higher precision is used. When overlap is detected (one object enters or rides on another), associated tracking is activated to maintain flexibility, thus optimizing the balance between precision and adaptability according to actual conditions
4Reliability
If continuous monitoring of all objects is performed, then reliability is maintained, but energy consumption increases
Solution Approach 1:
Instead of continuously performing full tracking processing on all objects regardless of their state, the system performs partial tracking only when necessary. When objects are separate, independent tracking is performed. When overlap is detected, associated tracking is performed only for the affected objects, reducing overall processing energy while maintaining tracking reliability through condition-based monitoring
Data Source
AI summary
An object tracking apparatus (2000) detects an object from video data (12) and performs tracking processing of the detected object. The object tracking apparatus (2000) detects that a predetermined condition is satisfied for a first object (20) and a second object (30), using the video data (12). Thereafter, the object tracking apparatus (2000) performs tracking processing of the second object (30) in association with the first object (20).


