3D Point Cloud Surround View for Vehicle Distortion Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Surround monitoring systems introduce significant distortion in virtual views due to the projection of images over a surface, leading to stretched or compressed portions that do not accurately represent reality.
Innovation Solution
A surround view system that generates a three-dimensional model of the vehicle's surrounding environment using spatial and temporal data from various sources, including sensors and external data, and maps visual data onto this model to reduce distortion and provide a more informative and efficient virtual surround view.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If images are projected over a surface to generate virtual views, then a surround view can be displayed to assist the driver, but significant distortion occurs causing stretched or compressed portions that do not accurately represent reality
Solution Approach 1:
The patent transitions from traditional 2D image projection onto curved surfaces to a 3D point cloud representation. By converting camera images into three-dimensional point clouds with depth information, the system eliminates the distortion inherent in projecting 2D images onto curved display surfaces. Each point in the cloud represents a spatial coordinate (x, y, z) with associated visual data, preserving geometric accuracy while enabling surround view generation.
Solution Approach 2:
The patent creates a virtual copy of the real-world environment by generating a three-dimensional point cloud model that replicates the spatial arrangement and visual characteristics of surrounding objects. This virtual 3D model serves as an accurate representation that can be manipulated and displayed without the distortion problems of traditional projection methods, allowing the driver to view the environment from multiple perspectives including bird's-eye views.
2Measurement precision
If traditional projection methods are used to generate surround views, then the system structure is simpler, but the visual quality and accuracy are significantly reduced due to distortion
Solution Approach 1:
The system enhances measurement precision by adding the third dimension (depth) to traditional 2D image processing. By generating three-dimensional point clouds from multiple camera views, the system achieves accurate spatial representation of surrounding objects. The point cloud data structure (x, y, z coordinates) inherently preserves geometric relationships and distances, providing superior spatial accuracy compared to projected 2D images.
Solution Approach 2:
The patent replaces the mechanical/optical projection system with a computational 3D modeling system. Instead of physically projecting images onto curved surfaces, the system uses algorithms to construct virtual 3D representations from camera data. This substitution of computational methods for physical projection eliminates the inherent distortion of projection while maintaining system feasibility through software-based processing.
Data Source
AI summary
A surround view system for a vehicle includes: a three-dimensional geometry unit configured to receive vehicle surrounding environment data and generate a three-dimensional model of a vehicle surrounding environment based on the received vehicle surrounding environment data; and a view generation unit configured to receive visual data and map the visual data to portions of the three-dimensional model based on the spatial and/or temporal information of the visual data and the vehicle surrounding environment data to form a virtual surround view.