Virtual LiDAR Sensor Using Camera Modules for Intensity Point Clouds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current motor vehicle sensor systems, particularly those using LiDAR technology for cruise control and driver assistance, require an enhanced method to evaluate surroundings effectively, as they lack comprehensive spatial and intensity information for improved perception.
Innovation Solution
A virtual LiDAR sensor system comprising multiple camera modules and algorithm modules that generate depth, RGB, and segmentation images, converting these into 3D point clouds with intensity information, utilizing 3D projection and incident angle modules to provide detailed spatial and intensity data for enhanced perception.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If current sensor systems with cameras and sensors are used for cruise control and driver assistance, then basic surrounding evaluation is achieved, but comprehensive spatial and intensity information for improved perception is lacking
Solution Approach 1:
The patent combines multiple camera modules (RGB camera, depth camera) to capture both color and depth information simultaneously. This merging of multiple sensing modalities creates a comprehensive point cloud that includes both spatial coordinates and intensity information, resolving the information loss problem while maintaining measurement precision
Solution Approach 2:
The system creates a virtual LiDAR point cloud that copies and enhances real-world spatial information by generating synthetic intensity values based on camera capture data. This virtual copying process reconstructs comprehensive spatial and intensity information that would otherwise be lost, achieving improved measurement precision without physical LiDAR hardware
2Measurement precision
If multiple camera modules and algorithm modules are used to generate depth, RGB, and segmentation images, then comprehensive spatial and intensity data is achieved, but system complexity increases
Solution Approach 1:
The system uses universal camera modules that can capture multiple types of information (RGB, depth, intensity) through different processing algorithms rather than requiring separate specialized sensors for each function. This multi-functionality approach achieves comprehensive spatial and intensity data while reducing overall device complexity compared to using multiple dedicated sensor types
Solution Approach 2:
The patent replaces complex physical LiDAR scanning mechanisms with a computational approach using standard camera modules and image processing algorithms. This substitution of mechanical LiDAR systems with optical-camera-based virtual LiDAR generation simplifies the physical system while maintaining measurement precision through software-based point cloud generation
Data Source
AI summary
A virtual LiDAR sensor system fora motor vehicle includes a plurality of camera modules and algorithms that generate a depth image, a RGB image, and a segmentation information image. The system is implemented with an algorithm that associates the backscattered signals with information from a color-reflectivity table, incident angle determination and depth information.