LiDAR Coordinate Calibration for Vehicle Road Surface Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing outside environment recognition devices face performance degradation due to changes in the installation posture of distance measuring sensors, especially when the sensor's posture deviates from the design value, and are limited to specific types of irradiation methods.
Innovation Solution
An outside environment recognition device that includes a storage unit for posture information, a coordinate signal conversion unit, a road surface candidate point extraction unit, a road surface plane estimation unit, and a calibration amount calculation unit, which converts observation points into three-dimensional coordinate signals and calculates calibration amounts based on a predetermined coordinate system, allowing for accurate recognition of the road surface and sensor posture regardless of the irradiation type.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the installation posture of the distance measuring sensor deviates from the design value, then the coordinate signal of the three-dimensional coordinates converted from the distance information does not coincide with the position of the actual object, but the recognition performance degradation can be suppressed by calculating calibration amounts
Solution Approach 1:
The patent changes the parameter of the coordinate system by calculating calibration amounts that transform coordinates from the sensor's coordinate system to the vehicle's coordinate system. This parameter transformation compensates for installation posture deviations and enables accurate coordinate conversion despite physical misalignment
Solution Approach 2:
The patent introduces an intermediary coordinate transformation process that acts as a mediator between the sensor's local coordinate system and the vehicle's global coordinate system. The calibration amount calculation serves as this intermediary mechanism, enabling accurate position recognition without requiring perfect physical alignment
2Productivity
If the outside environment recognition device uses the fact that the change in the distance of the observation point of the planar structure is small, then not only the road surface but also the observation point of the object having the planar structure such as the wall of a building or the side panel of the platform of a truck is adopted as the road surface candidate point, but the road surface plane estimated from the road surface candidate points includes many errors
Solution Approach 1:
The patent applies local quality by using different selection criteria for different spatial regions. It prioritizes candidate points within a specific height range from the reference plane (road surface level) while excluding points at higher elevations. This localized filtering ensures that only points likely to be actual road surface are selected, improving estimation accuracy without significantly reducing detection speed
3Adaptability or versatility
If the outside environment recognition device is based on a sweep irradiation type distance measuring sensor having a mechanical rotation mechanism, then it can estimate the road surface by sweeping and irradiating laser light, but it cannot be applied to other flash type distance measuring sensors
Solution Approach 1:
The patent achieves universality by designing a coordinate conversion method that works with any distance measuring sensor type (sweep irradiation, flash, or other types). The calibration amount calculation and coordinate transformation process are sensor-type-agnostic, enabling the same algorithm to be applied across different sensor technologies without requiring sensor-specific modifications
Solution Approach 2:
The patent replaces the mechanical rotation mechanism dependency with a computational approach. Instead of relying on physical sweeping motion to map the environment, the system uses mathematical coordinate transformations and calibration calculations that work regardless of how the laser light is irradiated, thereby eliminating the constraint of mechanical rotation mechanisms
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 solution effectively suppresses recognition performance degradation caused by changes in the sensor's installation posture and can be applied to various types of distance measuring sensors, ensuring accurate environmental recognition and notification of potential issues.
Implementation Method 1
The distance measuring sensor measures a distance to an object by emitting laser light in a pulse shape and measuring a time interval until reflected light from the object is received
Implementation Method 2
a distance measuring sensor such as a laser imaging detection and ranging (LiDAR) sensor
Data Source
AI summary
Regardless of the irradiation method of the distance measuring sensor, degradation of recognition performance due to a change in the installation position of the distance measuring sensor is suppressed. An outside environment recognition device (12) that recognizes an outside environment around a vehicle according to an observation result of a LiDAR sensor (11) installed in the vehicle is configured to include a storage unit (21) that stores posture information of an installation posture of the LiDAR sensor in a three-dimensional predetermined coordinate system, a coordinate signal conversion unit (23) that converts a plurality of observation points obtained from the LiDAR sensor into a plurality of three-dimensional coordinate signals on the basis of the posture information in the predetermined coordinate system, a road surface candidate point extraction unit (24) that extracts a plurality of road surface candidate points indicating a road surface from the plurality of three-dimensional coordinate signals based on a height component of each of the three-dimensional coordinate signals, a road surface plane estimation unit (25) that estimates a road surface plane based on the plurality of road surface candidate points, and a calibration amount calculation unit (26) that calculates a calibration amount of the posture information based on a reference plane set based on the predetermined coordinate system and the road surface plane.