Simulated Point Cloud Generation via Virtual Scanner
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for generating point cloud data simulations are costly and labor-intensive, particularly when creating large-scale scenes with many obstacles, as they require manual 3D modeling or the use of high-precision laser scanners that are inconvenient to transport.
Innovation Solution
A method that acquires point cloud data from an actual environment without dynamic obstacles, sets dynamic obstacles in a coordinate system, and simulates scanning lights to update the data, generating simulated point cloud data without the need for a 3D scene map, thereby reducing costs and complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual 3D modeling is used to create scene maps and obstacle models, then realistic scene simulation can be achieved, but production costs and labor requirements increase significantly
Solution Approach 1:
The patent uses real captured point cloud data of the static environment as a copy of the actual scene, replacing manual 3D modeling. This copy is then processed through coordinate system transformations and virtual scanner simulations to generate training data, eliminating the need for costly manual model creation while maintaining scene realism
Solution Approach 2:
The system uses automatically captured point cloud data from road collecting devices to generate scene maps without human intervention. The virtual scanner simulation and dynamic obstacle insertion processes are fully automated, requiring no manual 3D modeling work while producing realistic training scenarios
2Measurement precision
If high-precision laser scanners are used to acquire 3D environment data, then high-precision and high-density point cloud data can be obtained, but device portability and transportation convenience deteriorate
Solution Approach 1:
The patent uses point cloud data captured by road collecting devices (which are more portable) as a substitute for data from high-precision laser scanners. The captured data is then processed through coordinate system transformations and virtual scanner simulations to generate high-precision training data, achieving the desired precision without requiring portable high-precision scanning equipment
Solution Approach 2:
The patent replaces the mechanical high-precision laser scanner with a computational approach: using data from more portable road collecting devices and simulating the scanning process through virtual scanners in software. This substitution maintains measurement precision through computational methods while improving device portability
3Manufacturing precision
If manual or automated 3D scene map generation is performed, then accurate environment representation is achieved, but time consumption and processing complexity increase
Solution Approach 1:
The patent performs preliminary processing of captured point cloud data by establishing coordinate systems and identifying static environment features before generating the scene map. This preliminary action organizes the data in advance, making subsequent virtual scanner simulations and training data generation more efficient and reducing overall processing time
Solution Approach 2:
The patent uses captured point cloud data as a direct copy of the static environment, eliminating the time-consuming manual or automated 3D modeling process. This copy is then transformed through coordinate systems and virtual simulations to generate accurate environment representations more quickly than traditional methods
Data Source
AI summary
A method for generating simulated point cloud data, a device, and a storage medium includes: acquiring at least one frame of point cloud data collected by a road collecting device in an actual environment without a dynamic obstacle as static scene point cloud data; setting, least one dynamic obstacle in a coordinate system matching the static scene point cloud data; simulating in the coordinate system, a plurality of simulated scanning lights emitted by a virtual scanner located at an origin of the coordinate system; updating the static scene point cloud data according to intersections of the plurality of simulated scanning lights and the at least one dynamic obstacle to obtain the simulated point cloud data comprising point cloud data of the dynamic obstacle; and at least one of adding a set noise to the simulated point cloud data, and, deleting point cloud data corresponding to the dynamic obstacle according to a set ratio.