Vehicle Localization Using Stereo Camera and LIDAR
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current vehicle localization methods, such as GPS and Simultaneous Location and Mapping (SLAM), face limitations in precision and reliability, especially in parking lot environments due to signal blockage and the need for camera installation.
Innovation Solution
A method and system utilizing an entry stereo camera and LIDAR to capture and process images and distance data, constructing a 3D model of a vehicle entering a parking lot, and using parkway cameras to determine the vehicle's position through transformation matrices, ensuring high precision localization without reliance on external signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GPS is used for vehicle localization, then the system is simple to implement, but the precision degrades and fails when GPS signals are blocked
Solution Approach 1:
The patent introduces LIDAR as an intermediary measurement device between the vehicle and the environment. The LIDAR system emits laser beams and measures the reflected light to obtain precise distance measurements, serving as a mediator that enables accurate localization without relying on external GPS signals that may be blocked.
Solution Approach 2:
The patent replaces the electromagnetic wave-based GPS system with an optical-based LIDAR measurement system. By substituting radio wave detection with laser ranging, the system achieves reliable operation in environments where GPS signals are blocked, while maintaining or improving localization precision.
2Measurement precision
If Differential GPS is used for high precision localization, then the precision improves to 1 cm level, but the cost increases and it still suffers from signal blockage
Solution Approach 1:
The patent extracts the essential function of precise ranging from complex GPS systems and implements it through a simpler LIDAR-based approach. By taking out the core measurement function and implementing it with laser ranging and image processing, the system achieves comparable precision without the complexity and cost of Differential GPS infrastructure.
3Measurement precision
If SLAM with cameras is used for localization, then the precision reaches 20 cm level, but it requires mounting cameras on the vehicle
Solution Approach 1:
The patent creates a 3D copy or model of the vehicle using LIDAR scanning data. By constructing a digital representation of the vehicle's geometry and matching it with environmental features, the system achieves precise localization without requiring physical cameras to be mounted on the vehicle.
Solution Approach 2:
The patent transitions from 2D camera images to 3D LIDAR point cloud data for vehicle representation. By utilizing three-dimensional spatial information from LIDAR scans, the system achieves more accurate localization while eliminating the need for camera mounting, as the 3D data provides richer geometric constraints.
4Measurement precision
If LIDAR and stereo camera are used to construct 3D model, then the localization precision improves, but the processing complexity increases
Solution Approach 1:
The patent merges the 3D geometric information from LIDAR with the visual information from stereo cameras to create a unified vehicle model. By combining these two data sources, the system achieves higher precision localization while the integrated processing framework manages the complexity through coordinated use of both sensors.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach provides high precision vehicle localization within a parking lot, overcoming signal degradation issues and eliminating the need for camera installation, thereby enhancing reliability and accuracy.
Implementation Method 1
line scanning via at least one LIDAR an outline versus time of the vehicle entering the parking lot
Implementation Method 2
capturing via an entry stereo camera a template image and an entry video of a vehicle entering a parking lot
Data Source
AI summary
A method of high precision object location in a parking lot including capturing via an entry stereo camera a template image and an entry video of a vehicle entering a parking lot, measuring a distance versus time of the vehicle entering the parking lot based on the entry video, line scanning via at least one LIDAR an outline versus time of the vehicle entering the parking lot, constructing a scanned image based on the measured distance versus time and the outline versus time and forming a three dimensional construct of the vehicle based on the template image and the scanned image.