Virtual Image Display Using Multi-Angle Stereo Data Bases
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for generating and displaying three-dimensional data from stereo recordings fail to accurately represent vertical and inwardly inclined surfaces, leading to interpolation errors that distort the appearance of objects like house walls, making them appear inclined outwardly.
Innovation Solution
The method employs stereo recordings from at least three different angles to create data bases for texture and height information, combining data based on the display direction to accurately depict vertical and inclined surfaces, with optional additional recordings for enhanced image quality and interpolation between data bases.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If stereo recordings from only two positions are used, then the data storage and processing is simpler, but vertical surfaces and inwardly inclined surfaces cannot be accurately represented
Solution Approach 1:
The viewing area is divided into multiple solid angles, with each stereo recording covering a specific solid angle. Data bases are segmented according to these solid angles, allowing each recording to specialize in capturing surfaces visible from its particular viewpoint. This segmentation enables accurate representation of vertical surfaces by assigning the appropriate data base based on the viewing direction.
Solution Approach 2:
The solution transitions from two-dimensional horizontal coverage to three-dimensional angular coverage by introducing vertical elevation angles. Stereo recordings are taken from different elevation angles (e.g., 30 degrees from vertical), creating data bases that cover different solid angles. This dimensional expansion allows capture of vertical surfaces that are invisible from horizontal viewpoints.
2Manufacturing precision
If more stereo recordings are used to cover all directions, then image quality improves, but processing complexity increases
Solution Approach 1:
Data bases are pre-organized and pre-filtered according to solid angles and elevation angles during the data collection phase. When generating a virtual image, the system immediately identifies which pre-prepared data base corresponds to the desired viewing direction and retrieves it without requiring real-time processing of all recording data. This preliminary organization dramatically reduces processing complexity.
Solution Approach 2:
Different data bases are optimized for different local viewing conditions (specific solid angles and elevation angles). Each data base contains only the relevant texture and height information for its designated angular range. This local optimization allows the system to maintain high image quality for any viewing direction while keeping each individual data base manageable in size and processing requirements.
3Ease of manufacture
If interpolation is used to generate height information, then data processing is simplified, but vertical surfaces appear distorted and inclined outwardly
Solution Approach 1:
Instead of interpolating height information mathematically, the system creates actual stereo recordings that directly capture vertical surfaces from appropriate elevation angles. The texture and height data are copied directly from these recordings into dedicated data bases, preserving the true geometry of vertical surfaces without distortion or mathematical approximation.
Data Source
AI summary
A method for displaying a virtual image of three dimensional objects in an area using stereo recordings of the area for storing a pixel and a height for each point of the area. A method is obtained of enabling displaying of vertical surfaces or even slightly downwards and inwards inclined surfaces. Stereo recordings from at least three different stereo recordings of different solid angles are used. For each different solid angle at least one data base including data about texture and height pixel point wise is established. Data for displaying the virtual image are combined from the different data bases in dependence of the direction in which the virtual image is to be displayed.