Plane Detection Using Height Histograms for CGR Anchoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In augmented reality and virtual reality environments, placing computer-generated objects (CGR objects) in unmapped or dynamic scenes without suitable virtual substrates can result in poor user experiences, as these objects may float or collide with real-world objects, due to the lack of accurate detection of real-world planes.
Innovation Solution
The method involves generating plane hypotheses by obtaining a point cloud of the scene and creating a height histogram or line-space image to detect horizontal and vertical planes, allowing for the placement of CGR objects on real-world surfaces.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If plane detection is performed in unmapped or dynamic scenes, then CGR objects can be properly anchored to real-world surfaces, but the computational complexity and processing time increase
Solution Approach 1:
The plane detection process is segmented into distinct stages: point cloud acquisition, preprocessing/filtering, plane hypothesis generation, and validation. This segmentation allows each stage to be optimized independently and enables early termination when sufficient planes are detected, reducing overall computational complexity while maintaining detection accuracy.
Solution Approach 2:
The system performs preliminary filtering and processing of point cloud data before full plane detection is required. By pre-processing the point cloud and identifying obvious planar regions early, the system reduces the computational burden of subsequent detailed plane detection while ensuring accurate CGR object placement.
2Measurement precision
If comprehensive plane detection is performed to ensure accurate CGR object placement, then user experience improves, but processing time and computational resources increase
Solution Approach 1:
The system performs partial plane detection by detecting only the necessary number of planes required for the current scene and CGR object placement needs. Rather than detecting all possible planes in a scene, the system stops when sufficient planes are found, reducing processing time while maintaining adequate precision for the application.
Solution Approach 2:
The plane detection process is made dynamic and adaptive, adjusting the detection threshold and processing depth based on scene complexity and real-time performance requirements. This allows the system to achieve sufficient precision quickly in simple scenes while allocating more resources when needed for complex environments.
Data Source
AI summary
In one implementation, a method of generating a horizontal plane hypothesis includes obtaining a point cloud of a scene including a plurality of points in a gravity-aligned coordinate system. The method includes generating, based on the plurality of points, a height histogram indicative of a plurality of non-overlapping height ranges in the gravity-aligned coordinate system and a respective plurality of weights. The method includes generating one or more horizontal plane hypotheses based on the height histogram.