Hybrid Tracker Selection for Object Tracking Accuracy

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing tracking systems, such as feature-based and instance-based tracking systems, face challenges in accurately tracking objects in sequences of images with low quality or when objects lack distinct features, leading to loss of tracking or inefficiencies.

Innovation Solution

A hybrid tracking system that combines a feature-based tracking system and an instance-based tracking system, along with a tracker selection system, which compares confidence values to select the most accurate tracking data from each system, and includes a lightweight tracking system and sub-window tracking system to enhance accuracy and efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a feature-based tracking system is used, then tracking accuracy is improved when images have high quality and objects have sufficient features, but the system fails to track objects when images have low quality or objects lack sufficient features

Engineering Contradiction:

Improvetracking accuracyVSAvoidtracking reliability

Core Design Contradiction:

Measurement precisionVSReliability

Solution Approach 1:

The patent combines feature-based tracking and instance-based tracking into a hybrid tracking system. The feature-based tracker extracts features from image regions to track objects, while the instance-based tracker uses multiple instances of the object to improve tracking. By merging these two approaches, the system leverages the strengths of both methods to achieve reliable tracking across various image qualities and object characteristics.

Inventive Principle:

Principle #5Merging (Combining)

2Reliability

If an instance-based tracking system is used, then tracking reliability is improved when images have low quality or objects have few features, but the system fails to distinguish objects from similar regions when images contain multiple similar regions

Engineering Contradiction:

Improvetracking reliabilityVSAvoidtracking accuracy

Core Design Contradiction:

ReliabilityVSMeasurement precision

Solution Approach 1:

Inventive Principle:

Principle #5Merging (Combining)

3Reliability

If multiple tracking systems are combined into a hybrid system, then tracking reliability and accuracy are improved across different conditions, but system complexity increases

Engineering Contradiction:

Improvetracking reliabilityVSAvoidsystem complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent introduces a tracker selection system that acts as an intermediary between the feature-based tracker and instance-based tracker. This selection system receives tracking data and confidence values from both trackers, compares them, and selects the most reliable tracking data to output. This intermediary component manages the complexity of the hybrid system by providing a systematic method for integrating multiple trackers without requiring complex coordination mechanisms.

Inventive Principle:

Principle #24Intermediary (Mediator)

4Measurement precision

If a tracker selection system is implemented to compare confidence values, then tracking accuracy is improved by selecting the best tracking data, but processing time and computational resources increase

Engineering Contradiction:

Improvetracking accuracyVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The tracker selection system automatically compares confidence values from multiple trackers and selects the best tracking data without requiring external intervention or complex decision-making processes. Each tracker provides its own confidence value, and the selection system uses these self-provided metrics to make selection decisions, reducing the need for additional processing overhead.

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS9791541B2System and method to improve object tracking using multiple tracking systems

Publication Date: 2017.10.17 THE BOEING CO

AI summary

A method includes receiving first tracking data and a first confidence value at a tracker selection system from a first tracking system. The method includes receiving second tracking data and a second confidence value at the tracker selection system from a second tracking system. The tracking systems may be configured to track an object in a sequence of images, and the first and second tracking data may indicate locations of regions where the corresponding tracking system has tracked the object in an image of the sequence of images. The confidence values may indicate likelihoods that a corresponding tracking system is tracking the object. The method further includes providing output data to the first tracking system and to the second tracking system. The output data may include data selected from the first tracking data and the second tracking data based on a result of a comparison of the confidence values.