LiDAR Sensor Alignment via Image-LiDAR Object Orientation Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In autonomous vehicles, LiDAR sensor misalignment can degrade vehicle performance by reducing accuracy and efficiency, as complex systems require precise sensor alignment which is often difficult to maintain.
Innovation Solution
A LiDAR sensor alignment system comprising an imaging device, a LiDAR sensor, a mount device, and a controller that classifies objects in both image and LiDAR signals to confirm proper orientation and initiates adjustments to align the LiDAR sensor, using a processor and electronic storage medium to automate the alignment process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a large number of sensors are used in automated vehicles, then the system complexity and sensing capability increase, but the difficulty of maintaining sensor alignment and overall system reliability worsen
Solution Approach 1:
The patent implements a feedback mechanism where the imaging device continuously captures images of detected objects, the controller classifies objects and determines their expected orientations, compares expected versus actual orientations in LiDAR data, and generates adjustment commands to the mount device. This closed-loop feedback system automatically detects and corrects sensor misalignment, maintaining reliability despite system complexity.
Solution Approach 2:
The LiDAR sensor alignment system performs self-diagnosis and self-correction by using its own imaging device to monitor object orientations and automatically adjusting its sensor alignment through the mount device without external intervention. The system serves itself by detecting and correcting its own misalignment issues.
2Ease of manufacture
If manual sensor alignment methods are used, then the initial setup is simple, but the difficulty and time required for maintaining alignment worsens
Solution Approach 1:
The system performs preliminary alignment actions by pre-classifying objects and determining their expected orientations before LiDAR scanning. The controller prepares adjustment commands in advance based on image analysis, so when misalignment is detected, corrections can be made quickly without extensive real-time analysis.
Solution Approach 2:
The patent replaces manual mechanical alignment procedures with an automated electronic system. The controller electronically processes images and LiDAR data to detect misalignment, and the mount device uses motorized mechanisms instead of manual adjustment, substituting mechanical operations with automated control systems.
3Productivity
If sensor alignment is not corrected, then the system operation continues without interruption, but the accuracy and performance of the automated vehicle system degrades
Solution Approach 1:
The continuous feedback loop monitors sensor alignment in real-time during operation. The imaging device constantly captures images, the controller continuously compares expected versus actual object orientations, and adjustment commands are generated and executed automatically, maintaining precision without interrupting vehicle operations.
Solution Approach 2:
The alignment monitoring and correction process operates continuously during vehicle operation. The imaging device continuously images objects, the controller continuously processes data and detects misalignment, and the mount device continuously makes adjustments as needed, ensuring uninterrupted useful action while maintaining accuracy.
Data Source
AI summary
A Light Detection and Ranging (LiDAR) sensor alignment system includes an imaging device, a LiDAR sensor, a mount device, and a controller. The imaging device is configured to output an image signal associated with a first scene that includes an object. The LiDAR sensor is configured to output a LiDAR signal associated with a second scene. The mount device is attached to the LiDAR sensor and adapted to align the LiDAR sensor. The controller is configured to receive the image and LiDAR signals, classify the object from the first scene, and confirm the object is properly oriented within the second scene. If the object is not properly orientated within the second scene, the controller initiates an action.