User Tracking via Sensor Fusion and Probabilistic Estimation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing user detection and tracking systems face challenges due to variations in appearance, scale, rotation, position, and orientation, as well as factors like camera characteristics and illumination conditions, leading to inaccuracies and high computational costs.

Innovation Solution

Combining image-based tracking with inertial sensor-based motion detection, using sensor fusion techniques and probabilistic systems to aggregate data from multiple sensors, such as cameras and inertial sensors, to enhance tracking robustness and accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If image-based tracking is used alone, then tracking can be performed with visual data, but accuracy decreases under variations in appearance, scale, rotation, position, and orientation

Engineering Contradiction:
Improvetracking robustnessVSAvoiduser detection accuracy
Core Design Contradiction:
ReliabilityVSMeasurement precision

Solution Approach 1:

The patent combines image-based tracking with inertial sensor-based tracking to create a fused tracking system. The image processor generates head pose estimates from visual data while inertial sensors provide complementary motion data, and these are merged through sensor fusion algorithms to produce more accurate and robust tracking results that overcome the limitations of either method alone

Inventive Principle:
Principle #5Merging (Combining)

2Measurement precision

If multiple sensors and sensor fusion techniques are used, then tracking accuracy and robustness improve, but device complexity increases

Engineering Contradiction:
Improvetracking accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The tracking system is segmented into distinct functional modules: an image processing module that handles visual data, an inertial sensing module that handles sensor data, and a sensor fusion module that combines them. This modular segmentation allows each component to be optimized independently and simplifies the overall system architecture by clearly defining interfaces and responsibilities between components

Inventive Principle:
Principle #1Segmentation

3Measurement precision

If accurate user detection and tracking is implemented to handle variations in appearance, scale, rotation, position, and orientation, then detection precision improves, but computational cost increases

Engineering Contradiction:
Improvedetection precisionVSAvoidcomputational cost
Core Design Contradiction:
Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The system uses inertial sensors to provide continuous tracking data that partially compensates for the computational burden of image-based tracking. When visual data is sufficient, full image processing is performed; when visual data is degraded or unavailable, the system relies more heavily on inertial sensor data, performing only partial image processing or skipping it entirely, thus reducing computational cost while maintaining acceptable tracking precision

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentUS9832452B1Robust user detection and tracking
Publication Date: 2017.11.28 AMAZON TECH INC
  • US9832452B1 patent drawing
  • US9832452B1 patent drawing
  • US9832452B1 patent drawing

AI summary

Systems and approaches are provided for robustly detecting and tracking a user. Image data can be captured and processed to provide an estimated position and/or orientation of the user. Other sensor data, such as from an accelerometer and/or gyroscope, can be determined for a more robust estimation of the user's position and/or orientation. Multiple user detection processes and/or motion estimation approaches and their corresponding confidence levels can also be combined to determine a final estimated position and orientation of the user. The multiple user pose estimations and/or motion estimations can be combined via an approach such as probabilistic system modeling and maximum likelihood estimation.