Smart Glasses Pose Tracking via Image Masking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Inside-out tracking for smart glasses in a vehicle environment faces challenges in maintaining reliability due to false detections from moving surroundings objects and the need for accurate pose determination with low latency.
Innovation Solution
The method involves using a masking area in the camera image evaluation to exclude transparent panes and moving objects, allowing for pattern recognition only on fixed structures within the vehicle interior. This approach reduces computing power requirements and enhances tracking reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If inside-out tracking is used to determine smart glasses pose in a vehicle environment, then the system can operate with low latency and integrated pose detection, but false detections occur due to moving surroundings objects and transparent panes
Solution Approach 1:
The camera image is divided into a valid area and a masking area. The masking area contains transparent panes and moving objects that cause false detections. By segmenting the image evaluation into valid and invalid regions, the system maintains tracking reliability while preventing false detections from compromising pose determination accuracy.
Solution Approach 2:
The harmful elements (transparent panes and moving objects) are extracted from the image evaluation process by defining a masking area that excludes these regions. This extraction removes the source of false detections while preserving the valid tracking information in the unmasked areas.
2Measurement precision
If the entire camera image is evaluated for pattern recognition, then comprehensive pose information can be obtained, but computing power requirements increase
Solution Approach 1:
The camera image is segmented into a valid area for pattern recognition and a masking area to be excluded. This segmentation reduces the computational workload by limiting image evaluation to only the relevant portions, thereby decreasing computing power consumption while maintaining pose determination accuracy through focused analysis of valid regions.
Solution Approach 2:
Instead of evaluating the entire camera image, the system performs partial action by evaluating only the valid area outside the masking area. This partial evaluation is sufficient for accurate pose determination while significantly reducing the computational energy required compared to full-image processing.
Data Source
AI summary
A method for operating smart glasses for the augmented presentation of information objects on a display surface on the basis of a smart glasses pose in a vehicle interior is provided. The smart glasses pose is ascertained by inside-out tracking on at least one marker provided fixedly in a vehicle interior or at least one object provided in the vehicle interior. The at least one marker or the at least one object can be located by a computer-assisted pattern recognition method which is applied to a camera image of a smart glasses camera so that the smart glasses pose relative to the vehicle interior can be determined on the basis of a position, size and/or orientation of the at least one marker or of the at least one object. Only the camera image outside of a masking region is taken into account for the pattern recognition method.