Occlusion Mask Target Tracking for Mobile AR
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for real-time pose tracking in Augmented Reality (AR) on mobile devices face challenges in delivering six degrees of freedom with robustness and efficiency, particularly due to limited computational capabilities and the need for tight alignment with real-world objects, while also requiring simultaneous object detection and tracking within the camera frame duration.
Innovation Solution
A mobile platform employs an occlusion mask to prioritize target tracking, creating a mask indicating areas in the image to detect new targets, allowing for efficient detection by focusing on unmasked image areas and distributing detection over multiple frames, utilizing optimized computer vision techniques like FAST corner detectors and forest of spill trees for feature matching.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If object detection is performed across the entire image, then detection accuracy is improved, but processing time increases and real-time performance deteriorates
Solution Approach 1:
The patent divides the image into multiple regions of interest (ROIs) based on occlusion masks generated from tracked objects. Instead of processing the entire image, the detection algorithm focuses only on unmasked regions where new objects are likely to appear. This segmentation approach maintains detection accuracy while significantly reducing processing time and enabling real-time performance.
2Reliability
If tracking is performed with high priority to ensure real-time operation, then tracking quality is improved, but detection capability deteriorates due to limited time budget
Solution Approach 1:
The patent performs tracking operations first to generate occlusion masks that define regions of interest for detection. By completing tracking before detection within each frame, the system ensures high tracking quality while creating a structured approach for subsequent detection. The occlusion masks from tracking serve as preliminary guidance for where to focus detection efforts, enabling both functions to operate effectively within the time budget.
3Measurement precision
If the entire image is searched for new targets, then detection completeness is improved, but computational load increases beyond mobile device capabilities
Solution Approach 1:
The patent applies different processing strategies to different regions of the image. Regions covered by occlusion masks (where tracked objects are located) are excluded from detection, while unmasked regions are intensively searched. This local quality approach ensures detection completeness in relevant areas while significantly reducing computational load on mobile devices by avoiding processing in regions where new objects cannot appear.
Data Source
AI summary
A mobile platform detects and tracks at least one target in real-time, by tracking at least one target, and creating an occlusion mask indicating an area in a current image to detect a new target. The mobile platform searches the area of the current image indicated by the occlusion mask to detect the new target. The use of a mask to instruct the detection system where to look for new targets increases the speed of the detection task. Additionally, to achieve real-time operation, the detection and tracking is performed in the limited time budget of the (inter) frame duration. Tracking targets is given higher priority than detecting new targets. After tracking is completed, detection is performed in the remaining time budget for the frame duration. Detection for one frame, thus, may be performed over multiple frames.