Camera Object Tracking via Motion Sensor Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional camera tracking systems face difficulties in simultaneously tracking multiple objects due to increased computation requirements and fail to recognize objects that are occluded or outside the Region of Interest (ROI).
Innovation Solution
A camera apparatus equipped with both image sensors and motion sensors, which processes motion data to track object positions based on camera movement, reducing computational load and enabling continuous tracking even when objects are occluded or outside the ROI by alternating between vision-tracking and motion-tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional vision-tracking is used to track multiple target objects, then tracking capability is provided, but the amount of computation increases significantly
Solution Approach 1:
The patent divides tracking into two segments: vision-tracking for object identification and motion-tracking for continuous position estimation. This segmentation allows the system to use computationally intensive vision-tracking only when needed (for identification) and switch to computationally lighter motion-tracking for continuous monitoring, thereby reducing overall computation while maintaining multi-object tracking capability.
Solution Approach 2:
The patent dynamically switches between vision-tracking and motion-tracking based on operational needs. When an object is first detected or needs re-identification, vision-tracking is activated; once tracked, the system transitions to motion-tracking for continuous position estimation. This dynamic approach optimizes computational resources while maintaining tracking versatility.
2Measurement precision
If conventional vision-tracking is used, then object recognition is provided, but objects occluded by obstacles or outside ROI cannot be recognized
Solution Approach 1:
The patent merges vision-tracking (which provides accurate object recognition within ROI) with motion-tracking (which can estimate positions of occluded or out-of-ROI objects based on camera motion). By combining these two approaches, the system maintains recognition accuracy for visible objects while extending tracking coverage to include occluded and out-of-ROI objects through motion-based position estimation.
Solution Approach 2:
The patent uses camera motion data as an intermediary to bridge the gap between vision-tracking limitations and comprehensive tracking needs. Motion-tracking serves as a mediator that provides position estimates for objects that vision-tracking cannot detect (occluded or out-of-ROI objects), while vision-tracking continues to provide accurate recognition for visible objects. This intermediary approach resolves the contradiction between recognition accuracy and tracking coverage.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system efficiently tracks multiple objects with reduced computational burden and maintains tracking even when objects are occluded or outside the ROI, enhancing user convenience by presenting accurate object information.
Implementation Method 1
obtain, using the second sensor, motion data corresponding to movement of at least part of the apparatus identified in relation with obtaining the one or more images
Data Source
AI summary
Disclosed is a camera apparatus capable of tracking a target object based on motion of a camera sensed by a motion sensor and a method for tracking an object in the camera apparatus are provided. The method includes obtaining, by a electronic device including a first sensor and a second sensor, one or more images corresponding to at least one object using the first sensor, by displaying the one or more images via a display operatively coupled with the electronic device, obtaining motion data corresponding to movement of at least part of the electronic device identified in relation with obtaining the one or more images, tracking, from a corresponding displayed image of the one or more images, a position corresponding to the at least one object based at least in part on the motion data, and presenting, via the display, at least one portion of information with respect to the position corresponding to the at least one object.