Object Tracking Using Location-Specific Parameters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional object tracking algorithms struggle to accurately track objects in scenes with multiple objects moving at different speeds and directions, as they are optimized for specific types of movements and assume uniform behavior across the scene.
Innovation Solution
The method assigns location-specific and class-specific tracking parameters to objects within an image frame, allowing for different tracking values based on the object's position and class, enabling more accurate tracking of diverse movements by considering local statistics and uncertainties.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional object trackers use global tracking parameters optimized for specific movement patterns, then tracking accuracy for objects with predefined speeds and directions is improved, but tracking accuracy deteriorates for objects with different speeds and directions in the same scene
Solution Approach 1:
The patent applies local quality by dividing the image frame into multiple spatial zones and assigning different tracking parameters to each zone. This allows the system to track objects with different movement characteristics in different regions simultaneously, resolving the contradiction between optimized tracking for specific patterns and adaptability to diverse object types.
Solution Approach 2:
The patent segments the global tracking parameter set into multiple location-specific parameter sets. Each spatial zone has its own optimized parameters, enabling the system to maintain high tracking accuracy for different object types in different regions while using a unified tracking framework.
2Device complexity
If a single tracking parameter set is used for the entire scene, then device complexity is reduced, but tracking precision deteriorates for objects with varying movement characteristics
Solution Approach 1:
The patent introduces location-specific tracking parameters that vary across different spatial zones of the image frame. This allows the system to achieve high tracking precision for objects with different movement characteristics in different regions while maintaining a relatively simple overall architecture through automated parameter assignment based on object location.
Data Source
Figure 1A~1B
Figure 2A~2B
Figure 2C~2D
AI summary
A method (200) for tracking an object in a scene is provided, including receiving a first image frame (240a) of a first image stream (240) capturing a scene, detecting or tracking a first object in the first image frame, and determining that the first object is in a first part of the first image frame, determining, based on a at least one set (250a) of multiple values of a tracking parameter each specific for different parts of the first image frame, a first value of the tracking parameter for the first part of the first image, and tracking the first object in a second image frame of the first image stream subsequent to the first image frame, based on the first value of the tracking parameter. A corresponding device, camera monitoring system, computer program and computer program product are also provided.